An interesting article.
Minor quibbles:
1) there are more than three DVD decoders out there, and you failed to mention what is arguably the best (TheaterTek).
2) I'd have liked some more discussion abolut the relative benefits of using Windows Media Player or not.
3) On the first page:
If properly detected, the DVD player can reconstruct the original 24 fps source and eliminate the jaggies to deliver the original film's full frame of resolution.
Unless you have your Windows refresh rate set to an exact multiple of 24Hz, this is impossible. That's rather a shame - it would look a lot better refreshing at (say) 48 or 72Hz. With the refresh rate set to 60Hz, alternate frames will appear on the screen for different durations. Yes, the original frames can be reconstructed, but there will still be a lack of amoothness in motion, etc. because you aren't seeing the frames at the right times.
Large quibbles:
1) Why no mention of the difference between video- and film-deinterlacing? The whole 3:2 pulldown process shouldn't happen when the source is video rather than film. Do the players handle that correctly? And how, exactly, do they handle video-mode deinterlacing? Are they doing a simple "bob", or is it motion-adaptive, or region-based motion-adaptive, or per-pixel motion-adaptive...? Is there diagonal processing? All this makes a BIG difference.
2) I'm curious as to whether there's any difference in upscaling algorithms used if the video is played back at anything other than native resolution (e.g. upscaled to full screen). There might well be.
3) Worst of all, there is virtually no mention at all of 50Hz playback, which raises all kinds of problems.
First there's the refresh rate issue coming back to haunt you again - if the movie is 50 fields per second, how is that mapped onto a 60Hz display?
But even more seriously there's the issue of 2:2 pulldown - that is, of deinterlacing a 50Hz DVD. Detecting that you are viewing a source with 3:2 pulldown is relatively easy: certain sucessive fields will be exact duplicates. But distinguishing between film and video on 50Hz sources is extremely difficult - in fact it's a problem that has no rigorous solution.
So, all in all, you don't actually answer some of the most important questions. The two most important things to consider when reviewing video processing (on the grounds that they are the things most likely to be messed up) are: how well does it deinterlace video (rather than film)? and how good is it at distinguishing between film and video on 50Hz sources? Neither of these is really addressed in the article.