I have no idea about business or markets, if a game is good in its own right I just think its market will find it. I think most what are called AAA movies tend to just be about shiny special effects, a convincing trailer, but a poor movie. Not that I actually know what makes an AAA from a middle or small, but if the amount of advertising/hype is a factor then I'm basing it on that. I don't think I could stand to play another console game, there's some certain "sameness" that comes with them that reminds me of buffet food, where no matter which dish it is it somehow tastes like "buffet".
I think as a game or title becomes more popular (based on it's success), it's actually bad for it. It seems like the more millions they spend on a game, and the bigger the dev team, the more focus it loses. It turns out to be some giant committee build monument of compromises. I am of the opinion that too many games these days judge their own worth on how flashy they are, and how new their graphics are. On the other end of the scale are war games where it's almost like they go the other way, and try to impress their audience on how they can build an entire game using blue squares and red triangles. If Elemental was completely 2-d I wouldn't miss it at all. The only time I've actually rotated the camera was when I wanted to build on the other side of a mountain blocking my view.
Having played my share of Paradox games, I would like to say that although they have the right kind of idea in their depth, they seem to make simple tasks more complicated than they need be, forcing you to go through several menus for some basic things. And they tend to omit some simple features that you just have to be baffled how they could build an entire game and not (example:) allow you to change a units move orders after they have been set.
Looking at everything I just wrote, I find myself agreeing that a game should primarily try to be the game that as it was envisioned, and not worry so much about conforming to a genre, or a certain market. I think that's doing things oppositely from how it should be. An analyst can't build a classic, they can only give a formula for a middling title (sounds dangerously close to "middle-market"). I realize I haven't really made much comment on the original post, but I think digital distribution has yet to prove itself. And budget constraints should be taken into account before work on a title begins, when things get cut half-way through part of the vision is lost, and it shows in the final product.