Why would anyone suggest that?

Because I think you're untrustworthy.

Nat introduced an explicitly unreliable 18 month timeline on August 18th, 2000

That's all very convenient to your argument, retconning the P6 project manager as giving deliberately unreliable schedule advice (and justifying it that it's obviously unreliable here, 18 years later), just to save you from having to admit that Larry's original estimation was incredibly unrealistic.

I suspect Nat had gotten confused by this time.

Of course you do. You'll throw anyone who left the project under the bus if it prevents you from having to consider that anyone still involved with the project has ever made a mistake.

Would you care to explain "we expect to have alpha code a year from now" and the P6 team's repeated failure to give estimates longer than "18 months away" or "alpha code by next summer's YAPC" for several years?