I have also made the observation that coming back to mathematics after a few months makes it seem much easier. Several famous mathematicians pointed out this effect as well (I can only recall Bertrand Russell saying it at the moment). It even works on a day to day basis when it comes to solving problems.

In university, there were several times where I would spend all night working on some proofs and would go to sleep having made almost no progress. I would have dreams about the proofs (this made for restless sleep) but upon waking I would have many new insights into the problem and would be able to prove it. Conscious thinking about problems only seems to be reliable if you have faced (and solved) a similar problem before.

If there were some way to exploit this effect to maximum benefit, that would be awesome.