I think it's fair to say that this analysis applies better to web/app companies. There's just a fundamental problem with raising (and burning) millions trying to grow a site for inbound marketing, or an app for dating, or what have you, even though these are great ideas and things that people want:
You likely won't be able to generate real revenue on something whose value is so immediately replicable or replaceable.
There needs to be a caveat to "build something people want": there has to be some barrier so that others can't compete so easily. Customer and market understanding is one, and that's certainly helped. But again, the replication model, which will drive down profits for everyone, kicks in. Technical superiority is a possible solution, but most web/app products simply wouldn't benefit from this because their functionality is trivial.
Of course, there are exceptions. Unicorns. Things like Whatsapp, Twitter, and Salesforce come to mind. Obviously their initial products were replicable, but the winner-takes-all network effects (and eventual "platform dynamics") create that lock-in. For every one of them, 1000s of similar companies fail. The problem is when VC money is chasing those 1000s of likely-to-fail-I-might-do-it-too companies. They're more likely to fail than laundromats, but they're burning millions in an attempt to be those "winners". Easier than investing in clean energy though.
Fundamentally, as a society we need some technical advancement to create meaningful economic growth. We can't just keep advertising and selling the same stuff to each other. When a sector is running dry on innovation opportunities (i.e. webs/apps) maybe it's time to look elsewhere. Or stick around and wait for unicorns. Meh.
For now though, how people want to misappropriate (in my opinion) their money is their choice, as long as most of this doesn't get dumped on John Smith from The Public like it did in 2000.