[vintermann]

There have been men in their 20s marrying women in their 50s too (especially when inheritances have been involved) but since they wouldn't get children in that case, they would be totally invisible in these kinds of genetic statistics.

Even if average age in a relationship was symmetric, men being fertile for longer would cause age at (child's) conception to be higher for men. That effect needs to be untangled before we can know just how much men preferred younger women in the past/vice versa.