[vannevar]

There are a couple of obvious problems with the breadth-first approach. While I think everyone would agree that a great platform beats a bunch of great one-function apps, building a great one-function app is much cheaper and less risky than trying to build a platform. The number of people who know some narrow problem they can solve better than existing solutions is exponentially greater than the number of people who know how to build a broad platform that solves most problems better than existing solutions. And the reality is that most companies who pay for platforms will still buy one-function apps if the platform's functionality is inferior enough to replace that aspect. They are much less likely to buy another entire platform, however.

I doubt Parker actually personally believes what he's saying here, it seems more like marketing from a company that is trying to sell a platform than it does the thoughtful opinion of an individual.