When I thought (HA!) I was going to be the coolest guy around (Ha-Ha!)... I did so much reading on bi-amping last year... SMDH. *facepalm
So quoting a specific source isn't going to happen.
That said, when Bi-wiring and Bi-amping, they always specify that the wires going to any one speaker need to match as close as possible, as if it were a single wire instead of two. I remember one guy commenting that an inch didn't matter, but that feet could when going to the same speakers passive XO network. Same reason why they always seem to recommend using the same amps for passive bi-amping; a
just-in-case so you don't end up with a delay in the low or high signal line of the same speaker
if there is a difference in processing between two amps.
Yes, I understand a nanosecond here or there doesn't really matter.

And that's what... like 3' of speaker wire?
...But unlike Left speaker vs Right speaker, where the cable lengths aren't so important (within reason),every source I read said that Buy-bi-anything-ing

was potentially more stringent in tolerance (my word choice there).
(Yes, this was almost more than a passing fancy. I'm so VERY glad I chose not to pursue it, though!)
I know the rules and complexity change significantly when employing Active Bi-Amping, and this is where my limited Amplifier Education really shows. Beyond employing external Active XOs prior to the amp stage, I know people sometimes run lower-wattage Tubes for Highs and higher-wattage SS amps for Lows. I presume there is a way to address possible delay in different Amp stages at the active XO(?). Or maybe that just doesn't even com into play.
As always, if there is something I have wrong, I greatly appreciate the correction!!!