This seems a good message to reply to to say:

So how does it work?

I.e. what is the difference between MP+ and MPEG audio layer 3? Is it the same old ISO spec. with larger compression bitrates bolted on the top? Does it do something actually new?

The first half of this page seems to be an explanation of how MPEG layer 3 audio works, and the latter half has quite general and somewhat expansive claims. For instance "Lossless encoding using better Huffman codes" - which is basically technobabble as the Huffman coding is only done on the coefficients of the DCT - after all the 'unhearable' information is thrown out - and choosing a better Huffman code table would result in (slightly) increased compression, not lossless coding. Other claims on this and other pages are merely codec claims - 2x speed (on a pentium 3-750, surprise surprise), better quality for the same size, and so on. I haven't seen anything that tells me anything useful about the format itself, because there's nothing in the technical description that I can see that applies to only the MP+ format - although I may be missing it in all the jargon.

Anyone got a better grip on it?

Save the whales. Feed the hungry. Free the mallocs.