OS X also does sub-pixel-kerning. This is where the "format best for your display" comes into play. If you're using a sub-pixel display (with separate Red, Blue, and Green, instead of a single phosphor dot that could be any color) then they can provide improved kerning of black-and-white text. If you open a text document on an apple laptop, or attached directly to an apple display, and it is displaying a font with large diagonal lines, you will notice some color to it. If you type a "/" slash, and increase font size quite a bit so you can see it clearly and have a large white border to the left and right of it, you'll notice a distinct subtle color tint. One side of the slash will look a bit pink, and the other side will look a bit baby blue.
This is the effect of sub-pixel kerning at work. If you get out a magnifying glass you might be able to see it in action. Each square on an LCD display has three distinct parts to each square pixel. They're usually three vertical slices, one red, one green, and one blue. All three together make a square, so the sub-pixels are three times as tall as they are wide. If all three pixels are lit at once, they blend visually in your eye to form white, although you will still see the individual colors with a magnifying glass.
Green is almost always in the middle. If red is on the left, it's on the left for every pixel. So if you're going to draw a diagonal white line, you can either just light or dim all three RGB at once , OR you can "sub-pixel kern" by lighting only one or two of the sub-pixels in some rows. This makes the diagonal line a little smoother, and results in sharper looking text, especially at smaller point sizes. (this effect is much more pronounced with diagonal lines that are mostly vertical, making "W" one of the best text examples)
The color comes from the fact that if the left pixel is red, that means that in 2 of every 3 pixels on the left most side of the line, the red subpixel is OFF. (leaving either green or blue as the leftmost subpixel) On the right the opposite occurs, with blue being off 66% of the time. So the left side develops a green/blue tint, and the right side develops a green/red tint. These usually look baby blue on one side, and pink on the other side. The effect is very washed out and subtle, but can usually be noticed if you're looking for it. (and sorry, NOW you will probably not be able to avoid noticing it!)
Of course if the OS is unable to do it because it doesn't know the arrangement of the sub-pixels, or is unWILLING to support a monitor, you'll just get regular full-pixel kerning, and things will be a little more blocky looking. They can't just guess at this. I tried it - when decoding images, if you get the "endianness" wrong, this exact thing happens to the rendered image, but it's at the pixel level (usually in groups of 4 or 8 pixels), and it produces a very pronounced graphics issue. Seeing this, I played around with artificially creating this effect, and it produces seriously blurry text when the SPK is done backwards. (I was emulating a macintosh interface on a windows app, and was bothered by the much rougher looking text, and tried to fix it)
So yes, If your SPK kerning isn't working, it'll look less sharp. And again you won't be able to see this effect if you are using a non-Apple display even if you're using a macintosh computer. It has nothing to do with the monitor being able to do it, the OS just isn't bothering to try.
