One of the great things about software startups today is that it's very possible to reach “ramen profitability”. That's also one of the bad things. Once you get to “ramen profitability”, running out of cash is no longer a way to know that you should be starting afresh and trying something new. You can run a startup like that indefinitely — and many entrepreneurs will do just that, instead of building the next Dropbox and becoming legendary.
I'm of the opinion that people who are really driven to build great companies will use their "mediocre" companies as springboards to bigger and better things, and only those who don't care so much about how big their company is will remain "stuck in a dead end statup", so to speak (and they will be happy about it, because they will get their sense of fulfilment from somewhere else).
Building a startup that pays your salary and enables you to spend your time on whatever you want is no mean feat. Of course, if instead you build a startup that sucks all your time and energy in exchange for a meagre amount of money and success, then that's not a great deal, but then, surely, people stuck in this situation will not be any happier with it than they would be with a job that provided the same results, and they will act accordingly.
Count me firmly on the side that's very pleased that it's cheaper than ever to build a profitable startup.
If you read this far, you should follow my RSS feed here.