If I were you I would consider a model where I would be to do a full scan, display only the top X vulnerabilities found, and simply charge more to show the rest of the results.
Another thing I'm curious about: does this work on a pure client-side web application (e.g. my app is just one html page + javascript that loads all the html from templates)? Are you including static urls or somehow tracking "clicks" into a web app? Actually, most of the things I would concerned about in my web applications are things like not validating that I'm correctly doing correct validation in POST-requests. I'd be interested in seeing if you guys are doing that kind of "fuzzing" in that respect (though not sure if there is an automated way to do that safely). Additionally, I'd be curious to see what you detect Nessus/BurpSuite etc doesn't in terms of web application security.
Anyway, neat idea, perhaps I'll check it out a bit more thoroughly and do a comparison.