1) I think both whould be fine and have merit. Depends on how much you trust the Quality of your game. If its good enough (read excellent to classic) it might redeem the franchise. If its "just" good Elemental's "fame might drag it down.
I for one whould be happy if you whould give the game another month or 2 compared to the time announced on steam (those people wanting to play can always go for the beta...)
3) More content in the higher echelons (Tier 4 and especially 5 feels lacking regarding how hard especially Level 5 is to get no reason to not have 2 spells each at level 4 and 3 at Level 5 at least...) and combination Spells of Spheres 3-5. So really powerfull gone wild stuff we could drool over achieving (and Monsters could get...). As is Tier 1-3 mostly is actually more powerful at least in each-turn utility, possibly even overall. That seems off...
Global Spells whould be neat (something akin to the Global Spells of Master of Magic in Power, really I do mean it) if you could get those in time for release / beta 5.
Oh and for everyones sake: Make the escape spell either non-random or axe it alltogether (just keeping the Scrolls which are fine even if the spell is axed as a regular one).
It's a feature (imo a useful one and could be balanced by a considerable cost in Mana like 50 or 100) and sure fine as a mid-game tech (could well be later than its current place) but it shouldn't be random.
If you badly need random elemental spells do add something else for Air. Shouldn't be that hard to figure out something good. One of the worst decisions during the beta was to make escape 25% chance random imo. Retreat option with cost lowers necessary reloads while still costing and being tied to heroes.
Also don't forget to document the escape option and why it comes late in the game well. Otherwise its a sure receipe for confusion (since retreat is a rather expected feature in 4x. Don't expect understanding for randomizing that feature from 4x-veteran or greenhorn alike.)
Especially more Heroes gained from quests and more multi-tiered and difficult ones with good story. (The base of the game seems rather solid now. Content seems a bit lacking to me. Here's hoping you don't leave all of that to the modders...)
2-3 extra Wildlands whould also be neat and an option to controll how many and possibly even which Wildlands will be present in a given game.
To give you a little estimate of what i'm thinking here: 3-4 times the current diversity in quests whould be what I'd expect of content for a polished and finished releaseworth state for the quest system. Feel free to go beyond that. The more the merrier. (Especially mult-tiered Quests like the arena and simmilar. Less 1-Step Fed-Ex stuff please)
Please add in the return later option with quests. It shouldnt be do it or leave it alltogether (if accidently stepping on a tile for example...)
5) See 4.) more beef to the bones...
Alternatively: Bigger Impact but still scarce seems fine. Quest system has more beef / potential and covers most a random event system could offer so imo rather focus on that.
Another idea: Maybe use the random event system could be used to spawn new quest nodes from time to time to re-fill the world with adventure. (like Lairs all around the plane appear that must be cleared to get everyone xyz bonus or prevent a major planar desaster. Make it more interactive and dynamic. 1 time random without an incentive to react to it seems not that impressive to me to be honest.)
Master of Magic (and to a lesser degree Orion 2) did that rather well (but didn't have such an itricate quest system).
6) (riddiculous here)
Most Important thing: teach the AI to use strategic agressive Spells (Freeze, Tremor and Pillar of Flame I'm looking at you, they are some of the most deadly tools in the box of the seasoned player, low Levle to boot) and watch the players tremble. Wandering into AI influence should be a serious hazard.
Make the AI react to players strategic actions if somewhat possible (if an adversary fields fire/Cold immune Troops stop throwing fire/ice-mages at it. A player whould never do something that stupid and any AI beyond easy should refrain from something like that, same is true for other strategic experiences: spears if lots of swords are fielded and mounted troops against maces while Axes and Maces vs. Spears ect...)
Equip to counter adversaries favorite tactics. Player goes heavy on elemental stuff: Field troops with resistance gear, Lots of cutting weapons: Use Chainmail, Blunt: Use Plate ect.
Make the AI use tactical spells like blizard and fireball more often if lots of Stacks are fielded and make those spheres somewhat a priority if respective Shards are in plentiful supply for a given AI (same for Slow and Haste but reverse, hunting for those Shards if multiple mages are available with those spheres)
Make the AI actually use escape (be it from scrolls) when a fight goes south. (requires you take the suggestion under Spells about escape, badly needed change )
Make the AI actually pursue quests a bit more vigorously (if at all)
Make AI-heroes buy in their allies lands. No reason not to get that one handed lightning spear from your Kraxis friend. (While he still is your friend to later... ahem...)
Make the AI better at chosing city specialisations. currently most cities go the town route. Not sure if the odd conclave and especially more use of the odd fortress whouldn't help more (especially given how well the AI has become at producing large stacks of truly powerful units in high difficulties)
If an adversary in tactical is slow in intiative allready, use the slow spell on it (can win battles all by itself with enough water shards. Initiative below 10 usually means dead. 1 or 2 means sure death), use haste on units which pack a punch so they get to act more frequently (especially after engagement or shortly before it).
When lots of ranged troops are to fight against have the AI Army advance faster. Trying to get the first strike in fine and dandy. Leaves you vulnerable to projectiles though.
Make monsters ravenously eat unguarded or scarcely guarded pioneers. Settling in such a world shouldn't be a walk in the park (this should make players happy as well as making the game harder and more interesting since its equal opportunity beatdown). Having a bunch of defenseless blokes haplessly strolling around your backyard should lead to a full monster belly in short order. Intelligent or not. Attack 0 = fooder = Attack 100%.
When being two times or more faster in intiative while also having more moves: Move in, strike, move away. Don't just stand there and receive an unneccesary hit when you could get away without it. If webs or simmilar stuff is available as well as ranged projectiles: Try to hold powerful melee troops to get kills without much possible resistance. If that advantage is massive enough: No need to engage whatsover. Try to avoid contact then. Especially if initiative is truly superior, with or without haste.
Little comedic hint: Stop the souvereign leather leggear fetish sometime during polish or give a reason for it in the backstory. Getting 1000 Gold in sold leggear by turn 150 from a single vassalized enemy souvereign is the rule not the exception currently. No idea why the AI fanatically likes its leather greaves. Leaves me puzzled and rich every single time (AI should try that with healing potions, that whould beef up its power considerably while giving the player less of a reason to hit the sell button 200 times straight after getting an enemy to surrender)
Most of those things as visible are not intricate machiavellian plots you seem to look for but simple broad strokes that make the AI something else than braindead in the respective areas (strategic agressive spells and reaction to long-term strategic experiences are the most glaing ones).
Other fields the AI is quite formidable allready. Those stacks of powerful units boast quite impressive stats at riddiculous and are sometimes just beatable by tricks... (some hints: Freezing a stack and then peppering it with projectiles /spells while running away from the hopelessly slow pursuers or using tremor and then roasting an immobilized stack to crisp bacon with pillar of fire hurts... Countering slow with haste is also effective and cloak of nature is good for agreesively aiming for fire/cold/lightning immunity. A
iming for elemental immunity is very much worth it especially on already very powerful heroes / stacks. Pity to lose 12 perfectly powerful units to a single blizard or 2 when that spell is available or 5-6 to fireball / flame wave. Protection from fire / Cold make that even easier. When such a spell is incoming the respective protection should be a no-brainer as should counterspell if the chance is good but that last thing the AI actually manages to do at times...)
7) done: was willing to give an excellent this time. For a solid core though and very much bordering good still. Not least because I like those games and like rough diamonds. The game is not really release ready for the mass market yet or soon though imo. Lacks polish and more importantly broader content to become a classic or be even truly excellent. If you aim for excelent give it another month or 2 after the announced steam release date and meat up the solid bones. Don't let modders do so. Doesn't go down well with "the masses" even if the modders and seasoned 4x players might. Also games polished that extra month often very much look it (things like UI-polish truly pay off in "first impressions". Endless Space is a stellar example here what UI polish can positively do to a game)
8) good to hear. Very much looking forward to stellar mods a year or 2 down the line (if only time was more vast and life whouldn't get in the way of gaming...
10) looking very much forward to it and hope you'll acheive what you set out to do. For your sake as well. Hard work deserves respect.
12) Not a huge fan of campaigns but I'll give it a try nontheless and sing the praises if its stellar.
13) good to know, prefering client free but can live with both (do you consider offering it on GoG.com as well? Nice service and worth supporting)
14) sad to hear but very much understandable and not exactly a suprise. Probably a good choice.
15) 64 bit now on main rig and exited to see what you can offer for 64 bit in the future.