Yes, true. What I actually said was that if anyone turns us down simply due to the implicit valuation, they've failed an IQ test, but these things get compressed when reduced to one sentence quotes.
By the time you're at your YC interview, you already know all of the terms of the deal except for one variable: the percentage of the company taken by YC. So if you're accepted, PG calls and gives you this one piece of missing info, and you and your cofounders get 5 minutes to decide whether it's acceptable.
I can think of several situations where that 6% could get a better return than YCombinator:
If it's the difference between attracting a top-notch cofounder and a take-what-he-can-get cofounder. PG isn't going to write your code for you.
If it goes to an angel investor with even more experience and connections than YCombinator. I'd take Ram Shriram or Peter Thiel over Paul Graham any day...not that I could get them, but apparently I can't get PG either. ;-)
If it lets you hire key early employees. For an employee, the difference between 1% and 3% of equity is huge, and you'll be able to attract much more talented people if you give them more equity. A team of 5 "A" players will likely beat a team of 2 "A" players, 3 "B" players, and YCombinator.
I agree that it's silly to turn down YC just because the money/equity ratio is fairly low. But YC's not the be-all-and-end-all of seed funding: there are other resources out there where your equity may get better returns.
Not if you think someone else will improve your prospects by a bigger ratio compared to the cut they're taking though.
Don't know of anyone like that, just being pedantic.