Is There a Better Way to Calculate the Probability of Tennis Majors?

Ilo is a great system because you can essentially calculate it and recalculate it in your fingertips. It doesn't require a model and no maximum likelihood at calculation. But it also has some serious drawbacks. I'm curious to know whether there are some articles written about using more substantive statistical methods to adjust for some of those differences. And maybe that would be an interesting project if it hasn't been done.