Thanks to everyone for the replies. I guess I'll elaborate on what I said above and hopefully it'll be more clear on what I'm trying to accomplish and the reasoning behind I've made certain decisions so far.
First off, I've already written a program and a number of helper scripts to make an image of a CD. It's not perfect, but it's a start nonetheless. In the most basic form, I can insert a PCE game, run my script and it'll rip into iso/wav.
When I listed the goals (points A-D) in my original post, I left some details out. I'll try to explain the best I can. One big thing I left out was the process has to be cross-platform.
I figured I'd have to store the CD into some intermediate format (IFM). The flow would simply be original CD->IFM<->CD-R or Emu ISO Format. Some IFMs would be the commonly known ones such as iso/mp3, bin/cue, etc. These of course, would be saved to HD. Because my primary goal is archiving and restoring, any kind of lossy compression on the audio files isn't acceptable. Lossless is fine and I've been thinking of using FLAC to compress the audio.
zborgerd, thank you for your detailed reply. I did consider using cdrdao and cdrecord and I came to realize that having one big file just isn't practical. I'd like to queue up the GoT music in XMMS, Winamp, shoutcast it, or whatever and having one big image file just makes this cumbersome. That's why I'm making iso/wav "images" right now.
Also, although Daemon Tools is a kick ass program, it doesn't really help people who don't use Windows. However programs like Daemon Tools will facilitate the ability I mentioned in point A. Since Daemon Tools supports a large number of CD formats, one can write a script that makes, for example, a cuesheet from the IFM that Daemon Tools can use. In fact, I have scripts already that do this.
Because most people use Windows but yet I primarily use Linux and OSX, having everything cross-platform is very important to me. Unfortunately, my initial data track reading program only works in Linux. I'm trying to learn how to accomplish the same thing in Python so that it works in the major environments.
Finally, in order to protect the archive from becoming unusable due to tools no longer being available, I'm trying to do everything with open source programs.
In short, the intermediate format (for now) for the files is simply iso/wav. From there, I have scripts that will create a cuesheet that should work with any program that uses cue/iso/wav or toc file (cdrdao). I've been able to use my program to extract the Ys IV data track for use with the English patch and burn back to CD-R all in Linux so far. However, I haven't thoroughly tested the copy.
Hopefully I've explained enough so everyone knows what I'm attempting. If anybody has anything to contribute, please post. I still need to clean up the programs and add some more functionality. It's still alpha quality stuff. Once my new server is setup, I can post my programs if anyone wants to check them out. Just keep in mind, as of right now, it's Linux only.