I disagree with some of the comments made here. I don't think the study is that bad and I don't think Klimas has to justify herself.
The results aren't very surprising, perhaps rather obvious, but I prefer that researchers publish the data they have, especially if it's about a large group of 261 patients.
If I understand correctly, these were patients recruited for stress reduction trials. These patients had to fill in a CDC symptom questionnaire where PEM, amongst other symptoms, was assessed. So given the increasing interest for PEM, for example in the 2015 IOM report, the authors checked Fukuda CFS-patients with severe or very severe PEM differ from CFS patients who indicated very mild to moderate PEM.
There were some clear differences on the Fatigue Symptom Inventory and CDC CFS Inventory, but no significant difference on the Perceived Stress Scale or for Recent Negative Life Experiences. The authors focus on the results for mood and depression, which also showed a statistically significant difference between the two groups, although, as the authors admit, some of these might not survive corrections for multiple comparisons. I find it a bit strange that they didn't test what happens with these difference for mood and depression if they control for symptom severity.
If a ME/CFS patient with severe PEM has a depression than perhaps psychotherapy might be helpful. For me, it's mostly the sentence at the very end that bothered me, where they speculate that psychological distress associated with PEM is a maintaining factor by inhibiting patients from engaging in tasks.