I woukd use the 5E engine. Perhaps with a stretched bounded accuracy going to +6 to +11. Why reinvent the wheel? Everything else would be up for consideration.
Basically I woukdld survey for certain things. Hard decisions need to be made and some have to be done early. Eg game engine and complexity level.
I worked on my own edition of D&D during the D&D Next Playtest. I didn't have enough faith that they'd make a worthwhile game, and really liked some things from the playtest. I think the primary thing I would do is move back towards "Swingy" combat, rather than "slog" combat (where there's massive HP bloat).
Another big thing I want to see change is using BECMI's ability score modifiers, rather than the +/-1 per 2 ability modifier introducted in 3E. Probably means getting rid of ASI, but you can add direct modifiers without modifying the ability score itself. For example, something that normally gave +1 to strength might give +1 to strength checks and saving throws, but not attacks or DCs. Obviously there's a
lot of detail to work out for this.
Question 1. Complexity level.
This is self explanatory. B/X is a 1 or 2, 3.5 and 4E are an 8 or 9. Clearly explain the difference and pros and cons of both.
I'd probably like a 6 or 7. I like the fact that 5E doesn't try to have a rule for everything, allowing the DM to make rulings as needed. 3E tried to codify everything, and it made the game crawl when the DM had to look up a rule.
Question 2. Should D&D be a 10 level game (at launch). Magic would top out at 5th level spells. CR 20 would be the new toughest monsters. Otherwise 20 levels.
Uncapped levels, with limited benefits beyond 10th, similar to AD&D.
3. Archetypes. Should they stay or go? Partly related to 1.
Archetypes are always going to exist, but I'm assuming you're referring to subclasses from 5E. I think they're a great addition, allowing two characters of the same class without having lots of overlap.
4. Some spell effects would be revised or go away. Simulacrum and wish get revised. True polymorphism, shadechange go away.
Spells are going to be revised in a new edition, regardless. Very few spells need to go away, but many need adjustment.
5. Hit point and damage totals. Point out that 5.5 characters might deal lots of damage but monsters have bloated HP.
HP bloat has been a problem since 3E, and has just gotten worse with every edition. Having Con add HP using the +/-1 per 2 ability scores was a terrible idea. 4E was better since it capped the benefit to once, rather than every level, but it still massively bloated HP.
6. Poll for nasty effects using energy drain as an example. I wouldnt bring back old school energy drain but using exhaustion levels instead could be used.
I personally love negative effects, even ones that are "unfun" like paralyze. However, I agree that most negative effects should have limited impact on the duration of the character, and should have ways to be negated. Old school energy drain worked for a time when you had different level characters in a group by default, but losing levels in a game like 3E became a death spiral (lower level characters would just die more often, causing more level loss). I think petrification and death are the only two "permanent" effects I'm a fan of, and both of which can be negated with powerful magic.
7. Saving throw revision. Poll on if players like the current set up or want better scaling saves. Also point out monsters get them as well. Use pre 3E and 4E as examples. 3E and 5E are the odd ones out here.
It really depends on what kind of game you want. I'm personally okay with 5E's bounded accuracy making saving throws largely passable by everyone. The only change I'd consider is allowing half proficiency for "untrained" saves. Well, and completely rewrite the common usage so that every ability score was useful for a saving throw. This would make having a low ability score a significant penalty for about 1/6th the saves.
8. Potential overhaul of defenses vs magic. Expanded use of greater magic resistance and landing severe debuffs might require debuffing opponents first.
I'm personally a fan of the old percentage magic resistance, where some creatures were potentially immune to spells (making martials matter a lot more). Personally, I'd like a "shake it off" ability that allows a Legendary creature to spend some number of legendary actions to remove debuffs, but it can't happen at the end of the turn it's created (allowing at least 1 turn where the debuff is in effect).
9. Overhall damage dealing spells. Potentially go back to 3.5 suto scaling damage dealing spells. Or upcasting +2d6 vs 1d6 damage. This will overlap with 5. If players vote for less hp spell damage may be reduced as well. Basically look at buffing damage and nerfing save or sucks with 8 and 9.
I think the concept of spell damage needs a massive overhaul, but not for the same reason. First, I think that scaling cantrips need to go away. Using a spell slot should be a limited resource that does impressive damage. If you reduce HP bloat, then you don't need to upgrade the damage dealt. If you don't really reduce the bloat, then yeah spells need to add multiple dice for upcasting. Dice based on caster level just leads to the linear fighter/quadratic wizard problem.
10. Be clear designers can veto a survey. But mostly if its an impossible or contradictory results.
This already happens. They changed Stealth to become Invisibility during the One D&D Playtest without any input from the players. They do the survey to get an overall feel of how people think about things, but sometimes they've already made up their minds.