agreed, that was the obvious choice; when I was posting their front page was taking f-o-r-e-v-e-r to load, plus wanted to add some color commentary via 37signals.
I see in the blog that it's supposed to do RSS (I had a whole post typed up about how they should before I noticed) but I have a rather dumb idea that someone might like out there: a tape could be able to be an RSS subscriber as well; that is, have the twelve newest songs of a podcast cached and available on this website.
Also, though I don't know how (can you access ID3 information client-side? do hashing/audio fingerprinting?), you might be able to identify the file before it's fully uploaded, and just link in someone else's copy if it's already been uploaded.
take a look at kettle.js and the source code of a muxtape URL. you can download the files in the muxtape by grabbing the AWSaccessKeyID and Signature embedded in the HTML source (at the bottom) and building the URL for the source song from S3, e.g. http://muxtape.s3.amazonaws.com/songs/'+hex+'?PLEASE=DO_NOT_...
