There's a lot of this going on all around. But no game is perfect, and I think sometimes the true intentions from the devs get overthought and fail at some point and act the other way around. I think like this when thinking about indie games mostly, because when I'm thinking about AAA games it always comes to my mind the budgets and the fancy suit executives behind big companies that don't give a damn f*** about anything but the sale revenews and statistics, which in most cases is the only thing they really know about in the gaming busyness.
But I think some things must be mentioned about in reviews, as long as developers don't get too distracted by those things while deving. This is due to the diference this media has on other media types: Interactivity.
Gameplay is worth mentioning in a review, although it shouldn't be one of the affecting factors for the overall judgement. There are cases in which gameplay isn't very friendly but still doesn't affect the overal game experience at all (as a quick example, Both Call of Juarez games and Arx Fatalis felt kind of weird gameplay to me, but all are extremely recomendable by me).
I only made a review in my life, about
Assassins Creed, and I know I focused it on some technical aspects that ruined my own game experience, like some too-mechanical things, and voice acting, but I also focused on some important content aspects like repetitiveness.
I'm usualy a person who doesn't really believe reviews until I actualy experience the game. Most times when the review congratulates voice acting, I still find myself getting bored because of it while playing.
But I do the same on books and films. I recently recomended a book to a friend, saying that dialogues were a bit childish in some parts but the overall content was very worthy. And usualy I say that the actors in films totally suck for not looking like they are real persons in a real situation.
Also I often get happy when some one tells me the film I'm about to watch is 3 hours long, or the book I'm about to read is 500 pages long.
So from a consumer's point of view there's some important aspects are and should still be in reviews (Of course, knowing that models are very high poly and textures are very high res and physics engine is this one or that one, and that there's 50 ways to choose where you want to go or there's 5000 weapons to be collected, bla bla bla, is not at all the things I want to know when reading reviews... and it often happens...).
But yes, I agree many developers seem to dive into a swamp that they tend to not be able to get out for long while deving. Would that be because that's most of the feedback they end up getting from everywhere (including consumers)?
My review on AC was partialy made with that kind of belief in mind, and this gets more to the point you're trying to show. I was thinking they have lost at least 60% of their time making the cities beatiful and the movement/combat/climbing remarkable, but then they either forgot or didn't have any more time/money to make the game really interesting to play. The very same thought came to my mind when playing Oblivion - 4 years to make a world that imho is way too big and poorly functional and varied - Doom 3 - simple logical things like holding a flashlight in one hand and a pistol in the other are not possible. And very little time spend on how to scare the player? It's always 100% dark, which isn't scary, just frustrating. - and Bioshock too - Repetitiveness, immersion breaking non-deadly system... felt like playing Super Mario with the 1st level tiles repeated over and over forever, and with infinite Lives/Continues.