People balk at Lean Startup because it tends to come off as patronizing, reductionist, and antithetical to innovation. The stereotypical "lean" startup is so tethered to short-term customer feedback that it will produce, at best, an incrementally improved version of whatever product customers were already using. In other words, aiming for what customers will want a 2 months from now, as opposed to 2 years from now.
To some extent I think this criticism is justified, but it's important to remember that even if big bets and profound insights into what customers want turn out to be correct, "non-lean" startups still need to be agile enough to adjust what they're doing on the margins through fast experiments. Getting 95% of the way towards having a product someone will pay for just is not good enough. To use Color Labs as an example, it can also be argued that Color was REALLY close to being successful, they just messed up distribution, and that had they used a small-scale, more iterative roll-out approach things would have been different.
I like the idea behind the post (your startup idea does not come from the mouth of God, a fallacy many have had including yours truly), but it's basically just saying "do Lean Startup instead of a big launch." Why not both?
It's possible to run "in Stealth" and "Lean" and do a number of key learnings and Product/Market Fit tests ahead of time, then come out of "Stealth" with a big launch. (Case in point is Betable, where our Sheila, our Head of PR, ran a great launch). I really do think you can have your cake and eat it too. It's possible to be a Lean Startup and have a big Launch. You just need to put a Chinese wall between your marketing ("off the radar") and your PR until Launch time.
Edit: It's worth noting that what Justin is advocating that startups not do, which is a Color-style massive launch before any tests are done, is a completely valid point.
The other part of the fallacy is thinking that nobody but you could have thought of the idea you're thinking of, or conversely in our patent damaged world, that if you thought of it independently then you can't be infringing a patent because what you thought up was 'obvious.'
Both thoughts are wrong of course. If you're going to build a business around a concept these days it seems like you should always file a provisional patent if only to force the due diligence on patent infringement ahead of time.
To some extent I think this criticism is justified, but it's important to remember that even if big bets and profound insights into what customers want turn out to be correct, "non-lean" startups still need to be agile enough to adjust what they're doing on the margins through fast experiments. Getting 95% of the way towards having a product someone will pay for just is not good enough. To use Color Labs as an example, it can also be argued that Color was REALLY close to being successful, they just messed up distribution, and that had they used a small-scale, more iterative roll-out approach things would have been different.