But there were also unquestionable making a lot of comments about the size of the MMO world. WOW was HUGE at that point. (not that MMOs are not still huge, but it stood out as both dominant and very D&Dish).
This is correct - WoW was much bigger back then. And MMOs in general were - population-wise they're usually smaller now (WoW still peaks about the same size when a new expansion comes out, but its brief), and the really big difference is pop-culture. WoW hadn't reached it's pop-culture zenith when 4E was being developed, but it was certainly ascendant.
I agree too that your theories are largely compatible. I do think there was an idea that a new version of D&D that felt modern and exciting, and worked really well on a VTT or in organised play, could piggy-back on the general perceived popularity of video-game RPGs, both MMORPGs and Bioware-type games, to encourage people to come and try these previously "pen and paper" RPGs in a new format.
Organised Play as the same stuff as AL now was big, really big, back then, with 3.XE. At least it seemed to be to me - it was talked about more than AL is now, and I knew more people, online and off, whose primary mode of play was that. Also there were more FLGSes and the like back then to host it.
4E was very consciously designed in a way that would work well on a VTT. Core mechanics wouldn't work well in an MMORPG, because it's highly random, movement-centric and turn-based (neither translates well, as the Neverwinter MMO, which is based on 4E, shows - it might as well be based on 5E or 3E in a lot of ways). But for a VTT? Absolutely ideal. Stuff like the extremely complex initiative-fiddling and so on? That would work really well in a fully-integrated VTT. And we know that was in production from early on with 4E (before launch). It just didn't work out because of the tragedy then because of the studio they picked after that making decisions that ensured it couldn't work out.
The CCG inspiration is part of that of course, in that the abilities are very clear in how they operate, which is ideal for a VTT, because you could potentially do absolutely full integration with minimum hassle, and with minimum "asking the DM". Instead of player saying "I want to do this", they could have simply chosen the power they want, and clicked where they wanted it to go.
The only thing that conflicts with VTT stuff (and also would have conflicted with any kind of videogame) is the heavy post-L11 focus on interrupts and reactions and immediate actions. That's much harder to fully integrate, and suggests me that maybe they weren't totally married to the idea. I mean, if I was making a video game of 4E, as a turn-based tactics game, using the actual rules as much as possible, I'd have to remove a lot of the immediate action (the ones used out-of-turn) and interrupt abilities, because you just couldn't code that in in an attractive way (even reactions are problematic, because they break gameplay flow if not automated like they are in say, XCOM, but as they have a cost, and you don't pre-set them, you can't really fully automate them). So I'm not sure about that. It would work better in a VTT at least than a game (but works best in a real-world game).