Perhaps true enough with the mics and all in the electronic/rock world.
What I've been doing, it's the jazz world, so I'm mostly using mostly LD condenser mics and tightly set directional condensers. Sometimes there's a direct in from an electronic keyboard or an electric guitar,... I'm supplying no sound to the room. It's more like an acoustic/electric hybrid. I choose to use 24/96. It works for me.
Sets are rarely anything more than 45 minutes, but I always have all the sets pre-set as different projects so I can quickly move into a second or third set by changing projects. If I'm concerned, during an an extended applause after a number, I can also hit stop and then record again. Now I'm back to zero time for the file.
I do not worry about latency since I'm only doing audio. And no plugs to contend with, no processor load with that. Besides, why would you really need to? It's about getting the track down.
It's really about how much bandwidth can you push through the buss into your hard drive. It doesn't take a lot of proc for that but does require two hard drives and a reasonable amount of ram (2gig for xp 32 bit, 4 for 64 bit) and a machine that's tweaked for audio. (Davii: get the second hard drive! Your system might be getting bogged down trying to read and write at the same time. This is DAW 101 Shoot, borrow one for a test if the money's an issue.)
Again, hard drive is cheap these days, and the projects are easily moved onto DVD for back ups and to clear the space out for new projects,
The projects are then transferred to my studio computer where I do all the post live work. The external drive makes this easy. ( and I usually have a better chair to sit in!)
And as I said earlier I'm using a Pentium M for this. It's 6 years old but it's tweaked only for audio, has an external USB drive and 2 gig of ram. I'm also using Motu heads for the 16 channels I get. Even using the Nvidea firewire, go figure.
For what I do I like to do it this way.
It works for me,...