From the "relevant link":
"Participants' conformity was measured by how much they wished to conform to social expectations and be seen in a positive light, known as 'social desirability'. They completed a standardised measure and were also asked about their attitude towards paying tax. People who score highly on social desirability are more likely to conform, for example by paying tax, and agree with others. The researchers expected that they would be more likely to co-operate as well."
Properly, they are not measuring conformity. What they have done is asked people to signal whether they want to be identified as a "member in good standing". This seems like an excellent way to generate an error because it also selects for people who have an incentive to be seen this way. People who are not particularly good at working with others or who lack competence will need to signal group identity more. This could explain why "conformists" appear to be worse at cooperating.
This actually makes a lot of sense from Robin's signalling and status perspective. Those that conform to the socially expected norms need to differentiate themselves somehow, which means they are competing with each other on other terms. Those that don't conform in some ways need to show that they support the social group in other ways.
Hm. If conformists cooperated more, could one explain that situation using the same framework?
Conformists label themselves as going with the group, so they act in a way to fit the label. Nonconformists would do the opposite.
If I hadn't seen this post and I was presented with these two ideas, I would chose this explanation as more likely.
The great problem of status based theories is they are too good at explaining most possible outcomes when it comes to dealing with humans.
First read about this on Para Pundit, relevant link here.