The problem, of course, is that business accounting is complicated. And you can't easily pick a smaller sub-set of features to launch with (what good is invoicing without a chart of accounts, and what good is a chart of accounts without reconciliation, and you've also gotta be able to import the most popular file formats for bank data imports and such). It would take several people working pretty hard six months to a year to launch the smallest usable feature set, and they'd have to move fast to get to the point where bigger businesses could use it--revenues from very small businesses won't sustain this kind of company.

That's not to say it's not a great business idea. Between shutting down my previous company and starting my current one, I toyed with the idea pretty seriously. I may come back to it in a few years if the need has not yet been met and I'm ready to move on from what I'm doing now.