We didn't want token-based auth (i.e, Oauth2 or openstack), nor did we want shared-secret based auth (i.e., hmac). We specifically wanted to offer authentication with public-key crypto, and as the blog post points out, the usability of client-auth SSL sucks.

So when you say it's not solving a problem that hasn't already been solved, if you're talking about "the ability to sign a request" in the most generic sense, sure, it's already been solved N times.

Generally speaking though, in most of those schemes key management is an afterthought, and we didn't want it to be an afterthought. It's been solved by SSH, and that's what we wanted to leverage.; the point was not to reinvent the wheel, the point was to leverage an existing security mechanism.