The short answer is yes, you can. The catch is in how you are defining and measuring snow cover.
If it is approached categorically (e.g. presence/absence), look at how land-cover is used in other studies. Be sure to read the MaxEnt documentation on using categorical variables. If you are using measured snow cover across an extent rather than presence, you could approach it just like precipitation. The values of snow depth at each cell location.
Doing what you described so far, across a global extent, is going to take some processing time, depending on the data resolution.
Cass
USC SSI
[Not Dr. Kass!]