The thing about Julians' opinion is that he's going to be pretty much bang on the money when he grades the bat. He'll hit the bat and based on the sound and rebound he gets from his tapping up mallet he grades the bat.
We all do similar rebound tests be that with a mallet or bouncing a ball on the bat. We form our opinion which would generally be the same for a good bat and bad for a bad bat.
Now back to Julian but for everyone and take it one stage further to the manufacturering level. We drop a [predefined] ball from one metre and measure the rebound off of the bat. The height of that rebound determines the Grade of the bat.
Now that said and maybe you think it is a good idea but it isn't, there is an element of trust and consistency involved here. You may bounce the ball at the right spot on the bat, even on a stain and "ping" she bounces high G1++. You may have a bat that will come to life when you knock it in for bit G2 / G3 but you know it will be great.
The thing about this question is that it, as SillyShilly said, almost impossible to answer with a valid scenario as there are counter arguments for all types of grading, other then Julians that is :-)
edit: My opinion is that, if you can, grade on performance but do it honestly, cosmetically it won't be to far out either.