Why is the Mersenne twister prefered to using the wichmann hill algorithm to produce pseudo random numbers?

I read that in version by Python 2.3 https://stackless.readthedocs.io/en/2.7-slp/library/random.html the Mersenne twister replaced the Wichmann-Hill method as default generator. I was wondering if something can elaborate why this one would do that. I understand, that the period is a bigger for the Mersenne twister, but are there other features, that could drive this change?