The procedure worked great, I think, but now I have a conflict between the results from the polar alignment tool and PHD2's Guide Assistant.
I did not bother with parking; just used the HC to turn the mount to 4 hours east by the hour angle reading. I don't have any screen captures or logs, but I recall the measured polar misalignment was less than an arcminute. I've repeated this procedure several times, each time the error vector is very short (~ 10-20 pixels long), and it's difficult to make any improvements. With my image train I'm running at 0.63 arcsec/pixel, if that helps.
When I use PHD2's Guiding Assistant, it tells me the misalignment is ~ 7 arcmin. I have to wonder which measurement is correct. My guess is that PHD2 is more correct, since I can watch the DEC error slowly increase over time. If anything, maybe their scale factor is wrong, but I sort of doubt it.
Anyway, perhaps I should start a different thread on this, but I'd be interested to hear any advice. I'll get back with screen captures and logs next time I go out.