Cones have worked for some decades... where is the problem?
I liked the rule of three. I am confident, that multiclassing a la 3e can be fixed. I am sure that 4e hybridization could also work without extra efford, if class abilities are worded carefully.
So I am neutral about that.
Solo monsters shoul not be needed in D&D next, with the flat attack bonus. HP determine if they are solo or minion. Especially if high hp makes you immune to some effects.
Also I like the ability to play gridless, on a square and a hex grid. Makes everyone happy!
I also thought this was a really good Ro3, especially the directions for monsters.
With respect to solo monsters, based on the ideas about minions the other week I think we can safely assume they're going for a pre-4e approach of not needing separate designations for creatures, while using all the lessons gained from 4e to see what makes those kinds of fights work. Using hit points could really tie it together, just like you say.
For example, suppose something like "stun" has a different effect depending on if the creature has 0-1/4, 1/4-1/2, 1/2-3/4, 3/4-1 of its normal hit points. At the high levels the effect can be shaken off fairly easily, and maybe classic "stun-locking" isn't actually a thing until the 0-1/4 range. In this case the level 1 PCs facing a level 4 monster might find it just like a solo fight: the monster will have marginally (but not ridiculously) higher defenses and hit points, and the PCs will struggle to get that monster down to a point where the abilities that "break" solo fights actually work in that manner. When the PCs are 4th level themselves, they already do more damage, so those kinds of effects become effective much earlier in the fight. When the PCs are 8th level just a hit or two will be enough to make such attacks very effective. And when the PCs are 12th level that monster might be effectively minionized. The game didn't have to fuss over it, it is just something that happens smoothly, and conveniently dodges the sometimes confounding comparisons in 4e between, say, a level 7 solo and a level 14 standard.
To me the most interesting thought in the multiclass question is actually the part that involves the action economy. We've seen some hints in this direction already, but it seems the minor/move/standard action economy (a system I really liked) might go to something slightly less fixed, where if the game says you can take an action you can simply take it. What is really interesting here is that this sort of thinking can help support the solo monster issue as well: if an 8th level giant gets a 2nd attack, for example, it already has the extra effective (occasionally actual) actions that solos have in 4e with respect to PCs. In this case it comes from having a higher level, and again we can imagine a smooth progression from something like solo->elite->standard->minion without necessarily making those a separate thing. There are other problems with multiple attacks, particularly slower gameplay at higher levels, but as long as they stay away from the excesses of 3e (7 attacks, most of them misses!) I think they might actually be able to graft the best of 3e and 4e together. (And not just as lip service to 4e fans!) That would be quite an accomplishment! I've long been convinced that something like that is possible in principle, but when it comes to game design we need "constructive proofs" for it to be of any use, and this Ro3 convinces me they might actually be able to pull it off.
Are there other major aspects to solos that would need some love? I know MM3 revised how 4e approached solos, and I bailed on 4e before that. Still, given only the few paragraphs above I think we could handle:
Extra hp - Comes from higher level
Higher (but only slightly) defenses - Comes from higher level
Resilience against the classic lockdown effects - Comes from extra hp, and thus from higher level.
Extra actions (relative to PCs) - Comes from higher level
Extra damage - Comes in many forms including extra actions, extra base damage, and other improvements due to, yes, higher level.
The trickiest part might be hit points, because if a solo creature is designed for a fight with 4 PCs, one might naively suspect that by the time the PCs reach that monster's level they should have roughly 4 times as much hp as they had earlier. One can only do that so many times before everyone has hundreds or thousands of hit points. The trick is probably to make sure that the balance between higher monster defenses and lower PC damage in a solo fight is such that the "solo" monster only actually needs about twice the hit points as an average party member. Rough estimation: suppose the average PC starts with 15 hp (roughly splitting the 3e/4e difference), and a canoncial solo is level +6. That would indicate hp doublings for PCs roughly by levels 7, 13, and 19. Or 15*2*2*2=120. Would you feel OK with 80 hp wizards and 160 hp barbarians at 20th level? It's certainly in my temperate zone. Alternately, if one kept 4e-like tiers over 30 levels, those same numbers would arise if +10 levels represented a solo. Thematically speaking, it feels natural to me if a creature a tier up represents a suitable solo encounter.