OK there are a lot of things to answer here:
The problem with editing something close-to-but-not-quite 24fps in Premiere is that you will run into timebase problems. Premiere measures a lot of preview events and keyframe effects based on a
timecode. This timecode's relation to the frame # is based on the
timebase you choose. Premiere has a 24fps timebase, which means when your video is 24fps and your timebase is 24, then they sync up and you don't run into problems.
When you edit with something that isn't 24fps and set your timebase to 24 (since it's the closest number) you wind up with events and previews which are not in sync with what they should look like on the timeline. This is why we recommend using the 24fps method and not anything 23.98 or whatnot.
Now what confuses me about what you're saying is that your audio is at "25000". That makes very little sense, as your audio has no FPS or whatnot, audio's measured in totally different increments.
If you mean you're sticking your WAV file into BeSweet and it says 25000 there, that's because that's the default value for that field (i.e. when you install the program that stick that # in there automatically) and it has nothing to do with how fast your audio is.
If you are editing from PAL source (which is in 25fps) you should disregard a lot of the information in the guide because it's written for NTSC video sources (America, Japan, Canada, etc) and not for PAL video sources (England, Western Europe). PAL video sources can be made progressive at 25fps by simply running Telecide() on your video source, and that's it. Then you can edit 25fps video in Premiere with a 25fps timebase with no problem.
Your profile says you're in South Africa, and so I googled for what standard your country uses and
found this which says South Africa uses PAL, so I'll assume you're using PAL stuff.
If you are using PAL stuff, do not worry about slowing down things to 23.976fps or whatever, since PAL is in 25 to begin with and all you have to do is Telecide() the stream in AVIsynth to make it progressive. Then edit in Premiere using the PAL presets.
As for your question about interleaving audio - the relationship of audio to time in a video file is not related. An audio stream is not bound to a video stream except by the way it's packetized and interleaved, and the way that gets interleaved is based upon the information specified in the AVI headers.
When you're speeding up a song to match up with a new video framerate, you're not changing the audiofile's framerate or anything. Those numbers in between are just there for convenience - what it does is it divides the two numbers, multiplies that factor by the total length of the audio file, and then stretches or shrinks the file to that length, which is very different than telling a video file to be 24fps instead of 23.976 fps, because there you're not stretching anything, you're literally telling each frame to be played closer together, as opposed to the audio where you're changing the whole waveform.
Hope that made sense, if you have more questions feel free to post.