Kevin et al,
I am working on a blog post this morning about how the WWDC show the need for Wireless Capacity, which will include my thoughts about how to solve this issue. I will either post on my Computerworld Blog (http://blogs.computerworld.com/haider), right here on CWNP.com, or on my personal blog (http://wifijedi.com)
Kevin, you mentioned that there is a difference between "coverage" and "useable bandwidth". I totally agree with that statement. Just to be clear, there were no issues with the Xirrus WLAN at the Microsoft Professional Developer's Conference (PDC) with either of those. In fact, we (I am employed by Xirrus) had over 3,000 simultaneous connections in the same room.
The numbers vary, but the best estimates I have seen was that there were approximately 500 APs and 1000 users at the Apple WWDC, so that is well within the limits of what Xirrus already proved possible at the MS PDC (we actually planned for 6,000 simultaneous connections in the same room). OBTW, Xirrus also provides the Wi-Fi for other events such as INTEROP.
I won't put the spoiler here, but I can tell you the answer isn't related to "airtime fairness" or "electronic beam-forming". ;-)