So, you make some good points in this, and in your follow-up post, but a couple things bugged me, so I thought I'd throw them out here.
1) GR:AW was apparently designed with the idea that one-hit kills do not make for someone trying to drag themselves through the given level to the next save point over and over again. Though I think they addressed the health scale in varying difficulty levels so that it gets easier to kill your character. As far as killing the tangos? Aimpoint and any weapon and I can put a tango down with a shot to the head. Occasionally, they get back up again, another shot, fine. I hit his helmet the first time around, I guess, but whatever. It's deadly enough.
3) You have to understand the current environment and modes of thought of at least the US Army. The brass are busy saying that the future of conflicts will be CQB (Close Quarters Battle), urban warfare and so on. This is not to say they've completely changed their traditional training and modes of thought concerning other theaters of battle, but I think what the GR:AW folks were trying to do was take that current trend of thought and run with it. No, not a lot of long sight lines in an urban environment, and if you want to snipe, your options are limited. The tangos get to find high ground and snipe YOU, but not the other way around. I would at the very least like a way to get to where the enemy snipers were positioned...if only to pick up sniper ammo.
4) BF2 is NOT GR. Period. America's Army is...AA. AA is more about individual accomplishments (Army of One) regardless of the touting of teamwork and the like. GR:AW is its own game, IMO, and not really an improved GR.
5) Games will always have issues with that sort of thing. In campaign style missions and a continuous storyline, one may have to single out the player's character as some sort of ubermensch in order to drive the story. What can we say?
6) GR:AW and R6 are similar in some respects, but not enough. Both concentrate more on CQB than anything else, for one thing. R6 is more about hostage rescue and so on than anything else, and that requires different tactics and weapons. Have you ever BEEN to Mexico City? It may not be the washed out tan that everyone's used to complaining about everywhere, but wow. In R6-3, the tight quarters are to be expected for the most part. Lockdown, in case you haven't been blessed with it, is a waste of time and money, IMHO - avoid it at all costs.
7) I'm right there with you. I despise the spray and pray spawn-camping monstrosities of the MP shooter world for anything other than maybe ten minutes of playtime once a year just so I can rest assured my intelligence isn't atrophied... Seriously, though, ROE is important - there is supposedly a way to use ROE that is called out in the actual game manual, but damned if I can make it work on cue.
All this said, no game will be perfect. Some are less not-perfect than others. Each have their pros and cons. GR:AW could be a great game. Part of the replay value of R6-3 (for me, anyway) was the fact that tangos were not always in the same positions I remembered them in the last time I ran the campaign, or even from replay to replay of the same mission within the campaign. I think that's an important piece of replay value - otherwise you get into a routine when playing. GR:AW was too eager to make a lot of use of new technology to show off the beauty of the game without really taking into account the replay value - [GR] was and still is replayable out the wazoo, especially what with several fan-modded campaigns. If we could do that with GR:AW, it might help. I tend to agree with you about fire teams, maximizing fire, etc. Many folks who play shooters, however, are more into the BF2 style of 'teamwork' - even when playing COOP games (SWAT 3&4, AA, R6-3, etc.), which is a crying shame, really. If I'm thinking from a 'realistic' point of view, I don't want to be teamed up with Rambo - he's going to get my butt killed.
Just my own observations from afar.