I really don't know with the digital stuff, although you could test it fairly easily if you had a short and a long cable and a USB or FW hard drive. Just plug it in and run a hard drive benchmark with each cable and see if data transfer rates drop with one or the other cable.
I know that it's important with analog cables like monitors, and I know that all digital specifications have maximum cable length specs, but I haven't done any testing to see if interference is a problem with long serial cable lengths.
Cat5 for ethernet should not matter at all unless you're running hundreds of feet of cable or having problems using older wiring for gigabit ethernet (or so I've heard).
Most clear plastic sheath usb cables I've seen appear to have shielding underneath the sheathing, so maybe they're all shielded. A really flexible one may have less shielding, but I don't know. Again, if you're actually having problems with long cables, then your only solution may be to buy a high quality cable. If you're not having problems, or if you can measure the same performance with short and long cables, then it's probably not a big deal.
I have heard of issues with cheap DVI monitor cables, but I don't remember what the real issue was.