Oy, the research in HoI III wasn't all the rage either, to be honest. The kind of a problem I have with the 'field experience' stuff is that it will force you into a war early - and what's more, it seems like it will be a good move to, for example, start a colonial war early to get some field experience, because you can't get the same result via research. Which makes kind of a sense, but on the other hand, it should logically lead to Italy having better equipment across the board than Germany, since they were fighting Ethiopia in 1936. Not to mention Russian tanks that were build without any reasonable input from field experience, yet were superior to other tanks of the era.
The research, at the same time, seems to be completely linear and what is worse - streamlined. If it will be year-bottlenecked, like in HoI 3, it will be again rush at January to get better tanks, then better infantry, than airplanes etc. (permute for your personal strategy). I would much rather see a little more complex tech tree, with actual technologies leading to improvements in equipment, not just generic 'better infantry' techs. Like, you know, 'riveted armor' then 'cast armor', or something like that, and some of those techs would grant you bonuses. Preferably those bonuses would be procedurally assigned at the beginning of each playthrough and hidden from the player before researching given tech.
Example: one playthrough 'riveted armor' might mean +2 to your tank's armor, and 'cast armor' +1; next time it would be +1 for riveted, and +2 for cast; or even 0 for riveted and +3 for cast. That way it would be harder to optimize research, and you would sometimes waste time and money on useless technology. You know, like in real life.
But maybe you are right, it's not as much 'dumbing down' as moving the focus elsewhere - field experience. Whether this turns out to be a good idea or a bad one remains to be seen.
On that matter, I would also like to see more detail on combat level. The current model of "two lines of brigades randomly targeting and shooting each other, with any advantages/disadvantages simply altering the shooting/"defending against shooting" rate" doesn't do well to simulate the effects of unit-types like rocket artillery and of combat mobility allowing attackers to focus fire defensive positions in a certain place rather than trying to attack them all at once.
That would be great. But probably also resources - consuming, as anything more complex would probably require more processing time... Just look at the turn times in Gary Grigsby's games.