Soham....here are the reasons for all your burning questions...
1. 1600 x 1200 is the perfect resolution for high end testing (unless multi GPU is being used) simply because games stutter at anything higher.. The objective of any test is relativity of performance, not to measure the minimum fps that a card can churn out.
2. Also try running a 19 inch monitor at 1600x1200. Sure the monitor supports it, but you'll need to disable most of the eyecandy in games like FEAR (by eye candy I'm talking AA and AF, HDR, soft shadows etc) simply because the card won't be able to handle them. Once again we test at settings which achieve the best compromise between fps and visuals which is the eternal dilemma faced by all gamers. A game looks better at 1280x1024 with 4x AA and 8x AF than at 1600x1200 with all these settings off. I hope you get my drift here....
3. HDR has been used for farcry, splinter cell and FEAR.... DOOM 3 doesn't support HDR with SM3. I hope you realise that SM3 is a direct X effect while DOOM 3 is Open GL
4. Widescreen - well most games that we tested don't support widescreen natively, if at all. Farcry does support widescreen, but well let me ask you, how many widescreen gamers exist among our audience? Yes the numbers are on the rise, with widescreen TFT's, but a major chunk of people still opt for CRT's (no widescreen there) or regular TFT's. Maybe a couple of years down the line widescreen resolutions would make more sense, when a majority of gamers and games are using them.
5. Quake 4 is a very CPU bound game. i.e. it doesn't really scale well. I hope you realise we don't just wake up one day and go "ooh lets test graphics cards this month", we tried at least 10 games before finalising on the 4-5 most feasible ones. To further clarify, our initial tests on Quake 4 got nearly similar scores across cards in similar categories (with the usual Nvidia - ati differences of course). Additionally DOOM 3, Quake 4 and Prey use the same DOOM 3 engine, which has been optimised on the latter two games.