Don't forget, several people have posted that UG's (Touringplan's) crowd estimator formula also takes into account the wait times at the MK's "Big Three" (Space Mountain, Splash Mountain & Big Thunder Mountain).
Space Mountain is scheduled to close for it's rehab on April 19, I believe. Once it's closed, I imagine we'll see a related increase in the wait times for Splash Mountain and Big Thunder Mountain - as the people who would normally be riding/waiting in line for Space Mountain will be dispersed throughout the MK and many would be riding/waiting in line for the 2 other mountains.
Of course, that only takes MK into account, but still, if it's part of their crowd estimator formula, it makes sense.
So it sounds like it's a combination of Disney underestimating the success of the 4/3 deal and adding shows/increasing park times at the last-minute. As a result, the UG (Touringplans) sees that what they predicted to be 6's and 7's were actually 8's, 9's and 10's. And they update the rest of their numbers accordingly.
I am also under the belief that May won't be nearly as bad as the new numbers indicate. I'm confident the combination of Spring Break and the success of the 4/3 deal is what has been driving the crazy crowds the past few weeks. Spring Break is the multiplication factor in that formula. Once you remove it (i.e. late April/May), the effects shouldn't be as staggering.
However, I do believe the low May numbers we were originally going off of probably were too low and needed to be updated. I agree with some of the previous posters who said the UG might now be overestimating these crowds a bit to try to "cover their butts" (even though, as someone mentioned, they can't predict WDW crowds correctly when Disney can't predict them correctly themselves, so it's really no fault of their own).
Bottom line, I'm expecting 6's, 7's and 8's during most of May. Maybe an occasional 9 on the weekend. Granted, I'm not a professional and the forumula I'm using is "historical information + gut reaction".