Comment by mlyle

His claim-- as best as I can read it-- is that B leaking a long-length route changes where traffic is routed, but not to B.

It's possible he's saying something else, but I can't figure out, and he hasn't clarified.