If there is that 'g'/unhappiness correlation, maybe the causality is: unhappiness->'g'. The overly happy, seeing less problems, get less problem solving practice, whereas a tendency to be analytical could boost 'g' over a lifetime, though perhaps not effective intelligence.

I wouldn't expect this to apply to most readers, who get particular pleasure from solving intelligent problems. Think general population.