My sense of Graham's thesis - and I'm reading this really loosely here - is that he's saying a name's not worth sticking with if obvious metrics suggest it's a bad choice. The most obvious metric of whether you have a good name or not is whether it's already been taken, especially by someone else with their own idea and access to 6 or 10 or 20 dollars per year or whatever the annual registration fee is. The space for available names (the "name space", if you will ;) is still large enough that being very stubborn about your startup's names is probably hazardous, and certainly not a sign of health.

Ignoring the minutiae of the rest of the post (and you can argue I'm really doing this quite intently), I think I agree with the principle. The name's not that important. If you're processing financial transactions you could be named PayPal, sure, but you could also be named Stripe or Square or BrainTree[0] or any of dozens of weird names that make no sense at face value (if these names are in fact deeply meaningful, I'd be curious to hear the stories... privately, or on Twitter, not necessarily here/now).

There are reasonable things to get stuck on - if you're committed to being a B2C product or service, and others are telling you your startup is more of a B2B idea, then you should wrestle with that. But you shouldn't be wrestling over the name for such a great deal of time. If Paul Graham tells you to change your name when you're pitching to him so you can get a .com, you probably shouldn't (he argues) protest all that much.

I don't think I hold his view as strongly as he does, but I'm neither a successful VC nor widely respected in Silicon Valley, so his argument about perception is admittedly a little self-fulfilling and it seems impossible to reject.

[0] the irony is not lost on me, now that I check, that BrainTree doesn't own braintree.com - although first glance suggests Square only has squareup.com, square.com redirects to it; I'd be interested to hear Graham's thoughts on that, and I think it would signal strongly whether I understood his argument correctly.