They noted that the improvement is MORE if the display is larger, and less if the display is smaller. But a 20% jump in picture quality just by using SVideo is a great thing.
There is no theoretical limit to the horizontal resolution of the luminance component of S-video, although the most common source material encountered in S-video (from a DVD player) is 720 Max Pixels Across (540 TV Lines @ 4:3 or 405 TVL @ 16:9).
The Y part of S-video is normally equal to the Y part of interlaced component video.
It is conceivable that an HDTV set top box could deliver 1920 MPA via S-video but I know of no examples.
Component video allows much greater horizontal chrominance resolution (finer, clearer, and more precise transitions from one color to another).
Chrominance (the C part) horizontal resolution for S-video is spec'ed at about 160 MPA and I believe the theoretical maximum is 185 MPA.
On a DVD the chrominance resolution is half the luminance, or 360 MPA. I believe that HDTV uses two formats (a) like DVD horizontal and vertical chrominance resolution each half the luminance or every 2x2 block of luminance pixels share the same hue, or (b) horizontal is 1/4 the luminance and vertical equals the luminance where every 4 luminance pixels in a row share the same hue.
Video hints: http://members.aol.com/ajaynejr
Be sure you realize he said component- not COMPOSITE. Some people get the two confused. Composite is on the lower quality end (single RCA style cable), followed by svideo (multipin connector) then component is highest (set of 3 RCA style cables).
There is more info on various connection types, including pictures, in the FAQ and PRIMER.