Hi,
I can't answer question 1, but can confirm that you can improve your predictions by creating your own pseudoabsences. I'm using a method called Target Group Sampling, described in the paper below. If you refer to the sections on Background data and Sampling bias that should help :)
Merow, C., Smith, M.J. and Silander, J.A., Jr (2013), A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography, 36: 1058-1069.
https://doi.org/10.1111/j.1600-0587.2013.07872.x
cheers,
Sarah