I just watched the V4 crosstalk talk by Dr. Johnson. It was extremely informative. I've never heard him talk, before. He's really quite a good teacher. I wasn't sure what to expect, as I find his book a little jumbled. I was pleasantly surprised, however.
His description of how most of the intra-package crosstalk is due to magnetic coupling, and not merely the voltage deltas caused by di/dt through a particular L nor capacitive coupling due to dv/dt, was quite a surprise. The demonstration of the relative directions of the voltage swings (between aggressor and victim) made this very clear and convincing.
One thing I would have loved to have seen were the same tests (or simulations) with drive strengths bigger than 4mA/fast. In our designs, we typically need a mixture of LVCMOS 12mA fast and SSTL2_II, with about 200 I/O toggling together -- which makes for very noisy/jittery V2 and V2-Pros.
Anyway, good job, Xilinx.
Bob