Probably because taxi companies are a dime-a-dozen across various cities and states, and the capital required for an engineering investment to pull something like this off just isn't there at any given shop. It's a technology problem that the existing industry faces, and because the existing competition is essentially a coalition of independent companies, they can't band together to build something like this without a joined effort (nigh impossible).

I was recently in Pittsburgh and saw a clunky touch-screen app stuck to the back of the passenger seat that sounded just like Uber, so it's clear that someone out there is working on building Uber-like dependability for your taxi. But I also remember talking to the cab-driver on that trip, and having him tell me that their 'dispatch' is still one person sitting in an office somewhere manually dialing and dispatching cabs to received calls. The status quo in the industry is just so ancient in so many ways. Here's to hoping they can start moving and respond though.

Probably because it would be expensive and risky to roll out something across a large, already existing network, but at the same time running an innovative "side project" may well be met by anxiety (and corresponding resistance) by those who aren't able to participate. It seems to me that both of these responses are fairly economically rational, given the corresponding risks.

Regulation could be useful here, if properly designed, by helping to alleviate the fears of those in the existing industry who are most likely to be directly affected by these useful innovations. Unfortunately, the allergic political response of some sectors of society to anything interventionist makes coming to reasonable arrangements quite difficult, so we just end up with something that isn't really that great for anyone.