I've heard or read of special processing techniques that make images look better @ screen res when starting from 10-24 MP captures - I've always assumed this to be some extra sharpening (what I usually do for the web, seems to help a little), but was wondering if anyone does anything more specific?
There isn't that much special processing needed when reducing resolution. The type of interpolation used to reduce resolution has very little impact on the image quality, once you avoid the very poor choices of (bi)linear and nearest neighbor. So, bicubic, which is standard on virtually all image processing tools, gives excellent results.
One thing that does sometimes help to retain a little bit of extra crispness to the image, so that it looks a bit more like the original, is to use an 'unsharp mask' filter. This filter detects edges and sharpens them, but ignores film-grain or sensor noise. Some care over the settings is required, but with a little experimentation, you should be able to find a good setting.
I use a program called Breezebrowser pro to organise and manage my digital photos. When producing reduced res copies for screen viewing, it will automatically use bicubic interpolation and unsharp masking during processing. Most high-end image packages like photoshop can be scripted to do the same.
Interpolation for to increase the resolution of an image for printing (to avoid visible, blocky pixels) is a far more difficult task - and there are dedicated software packages specifically designed for that. This is what is discussed at the link posted by CattyKit. Nothing at that link is relevant to what has been asked here.