Unfortunately this advice doesn't apply well to a lot of consumer, ad-supported startups , but it would apply well to this guy's "modest"(nobody's ever given me $250k for anything) success (GroupScript).
More of argument against consumer startups than against this principle.
I have done it on the enterprise side, though, and the experience wasn't all bad. I built a face recognition system to spot TV reporters on the air and auto-tweet. It was a "neat"-enough idea that I was able to get 3 TV stations to run pilot programs. In the end, even though it doubled traffic to the TV stations' live streams, it turns out they barely make any money that way and unflattering screen shots irritated the reporters. You can argue that I should have researched harder up front, but if 3 different top-15 TV groups representing 125+ stations were willing to try it out, it wasn't far off. It was close enough to a big score and took so little time (1 year) to iterate, it was worth the try.
http://en.wikipedia.org/wiki/Mark_Zuckerberg
To what are you referring?
Also to take an example from the article how would one accurately assess demand for a 10 language flash gaming site without building it?
Even those who form a team and raise money are doing this. It's not just the bootstrappers. For instance, I can name 3-4 people in my admittedly tiny sphere who had an idea, raised significant amounts of money, built a product, promptly failed, then pivoted multiple times before running out of cash.
It's as if starting a business has become so easy, that it's easier to just do it and see what happens rather than invest any kind of time and effort in researching the market first. This shoot-first-ask-questions-later attitude is spurred along by stories of huge multi-billion-dollar companies being born out of half-assed "so what now?" pivots. (See Twitter.)
In my experience, it's ok to do this once or twice because in your first at-bat, you don't know what you're doing anyway, so stopping to assess a market is just a waste of time. Put a year or two into something, get a feel for how it's done, and if you don't turn into Mark Zuckerberg (home run on first shot), you'll be better prepared to assess the validity of future projects. (That said, the best of my 4 businesses was the first, so there are exceptions.)
While I admire his entrepreneurial spirit, the author of this article (and his partners?) seem to have taken the "shoot first" philosophy a bit too far. 10 startups in 7 years seems extreme to me. The optics of it, from a potential investor's perspective, seem negative, too.