Well, one upside of your idea is it sounds very testable. If your model's a good one, we should be able to throw in measurements of all the different traits and behaviors you list into a factor analysis or a PCA and pull out a general summer-winter factor/axis. At the least, the things you label as 'summer' traits should all positively correlate with each other, as should the 'winter' traits.

Some of the data for doing this test are probably already out there for the less obscure traits like gender, extraversion, promiscuity, polygamy and risk aversion. I would guess your theory would imply that other Big Five personality traits would have summer-winter correlations as well, like conscientiousness with winter ('More planning and preparation. Stockpiling resources. Savings.')

Not sure how I'd test the evolutionary psychology predictions, though...

(Edit - I might have been primed with this idea by having just read this cross-country comparison of how personality traits relate to gender and short-term mating behavior! Could be a good start for anyone interested enough to try testing whpearson's theory.)