Along the lines of what others have said... since they were using frequency of sex to approximate the male partner's sexual valuation of the woman rather than measuring that directly somehow, what this really tells us is that if a marriage is otherwise committed, sex and satisfaction seem to accompany one another, and if it's otherwise rocky (in that that one party at least is not so committed) sex and dissatisfaction seem to accompany one another - for the women involved in the (rather small) studies, anyway.
Their ideas about why that might be are just ideas, it seems like, since they don't mention asking the women why they were more/less dissatisfied under their circumstances. I can think of enough alternative explanations that unless they monitored more confounding variables/took more direct measurements than they say they did, I'm not going to take the headline too literally. It seems overreaching.
And yeah, someone you're married to (and presumably have a history with) who isn't committed to you is a whole different animal than a "short term relationship", or can be, so the two situations should not be conflated in the conclusions proposed, though they appear to have been.
Note: This is based on a reading of the abstract only, I haven't found a non-paywalled version of the full text.