As soon as DTube came out I started my stopwatch waiting for DSound - you worked fast! Of course DTube could just have supported what you are doing directly by allowing an audio-only file upload. Or a file where the video portion is blank or fixed, perhaps with closed-caption subtitle stream only - that would be good for karaoke fans. And you could do the same - audio file with synchronised lyrics - I'm sure there is a standard for this.
Anyway, one thing I think you have missed is a way for artists to claim their content. This is equivalent of the Steemit community (and bot) suppression of content or users that are plagiarism. I'd feel much more inclined to police plagiarism (not counting fair use) if I knew that doing so allowed the bonafide content creators to collect their dues. We shouldn't like posting of copyright material as-is any more than we like people who just copy and paste popular content on Steemit hoping to ride the upvote love of our fellow Steemians. That's especially true if it is copied without attribution.
For instance will there be an bot on DSound? That's effectively what YouTube and others do, except it is a much, much more aggressive bot with total-kill teeth and no mercy (or understanding of fair-use).
Without some of these things, and without support of original content creators you're in store for a legal nightmare I think. But I still applaud your efforst - I think Steem has really shown we don't need centralized media platforms, although to be honest I think the 7-day limit is pretty unfair. I'd much rather see a staged compensation scheme such that content a year old that suddenly gets discovered later on in its life can still generate revenue for the OP and those that discover it.
RE: Introducing DSound: a decentralized sound platform using STEEM and IPFS