I like parts of this article more than others.
For example, while I share a certain sense of frustration that SV seems to spend a disproportionate amount of resources on mobifotosocialgames, smartphones and their downmarket progeny are an enormous quality-of-life improvement for broader America. There's virtually nothing a cash-poor-responsibility-rich person needs to do that having a smartphone does not make them better at.
Commute to work? Apply for benefits? Schedule appointments with doctors? Successfully keep those appointments? Keep tabs on kids while working/errands? A smart phone makes you better.
As an entrepreneur who makes stuff for people who are not the usual suspects, the challenges to doing this are numerous and largely not conducive to the VC-funded trajectory. You have to convince people to pay you money for your services, you have to be able to service those customers profitably, and the customers are disproportionately pathological. Most of the software they consume is getting written by AppAmaGooFaceSoft because they can underwrite it with their massive monopoly rents, subsidize the cost straight down to zero, and deal with the CS headaches attendant in serving poor people by sending them to a call tree / web app backended by a Markov chain backended by /dev/null.
There's also a bit of "every problem has a software solution" enthusiasm which is, well, a bit overstated. (Sometimes this isn't stated outright, but when you expect to do something meaningful on e.g. $500k in an angel round, you implicitly expect to do something meaningful 100% in software, because $500k isn't even enough to launch a single McDonalds in the real world.)
Many problems have an 80% software solution attached to a 20% "interface with the real world" mandatory bit. The marginal cost of that last bit, though, is hundreds, thousands, or tens of thousands of dollars per client served. Or it has highly non-trivial political/regulatory barriers to entry, like "Convince an incompetent, intransigent, and politically invulnerable agency to disemploy half their workforce, who are by the way mostly veterans, whose main professional competence is doing an important thing slowly, poorly, and expensively."