The idea was to bump the requirements at the same time, to reduce the number of times a host would need to update their older systems.
The decision not to change any queries was made early on, 3.3 will bring a few query changes, There are queries which have been wanting MySQL 5 optimizations for awhile now, ultimately, this is going to bring performance improvements to (currently) 98%+ of users, with the remaining 2% likely to never update to 3.2+ or (based on your statement that 5 is slower than 4.1) slightly slower performance than previously,
Also, The WordPress stats[1] indicate that 2.7% of installs were on MySQL 4.1 (with 0.2% on 4.0, which was last supported on WordPress 2.8) - MySQL 4.1.2 has been required since WordPress 2.9.