Their findings don't seem to match with what Stanford is seeing. According to Ron Davies the problem is in the serum, not the cells.
It's entirely possible that the results are compatible.
The authors use two justifications for assuming there is some fundamental difference in NK cells from CFS/ME patients - the supposed differences in TRPM3 SNPs (which they know don't actually exist) and their previous studies showing reduced NK cell activity in patient's cells.
Well the first doesn't really exist and the second is a result that is often not replicated. Some labs have found reduced NK activity, many have not.
So actually there is no particular reason to think that these results are inherent in the cells, much as the authors would like us to believe this.
It is quite possible that the cell differences are the result of their previous existence bathed in serum. They have now been removed from the serum but it may take time for the effect of the agent(s) in CFS/ME serum on the cells to subside.
This is also a potential source of uncontrolled variability in assays of isolated cells. Small differences in processing time (that unexpected phone call, a machine that is not available etc, etc) could have a big effect on the cellular activity.
Maybe this too is the source of the inconsistent NK cell activity studies. There could be significant though small and unrecognised differences in processing of different laboratories which make a big difference to the outcome.
In other studies on energy production by PBMC from CFS/ME patients compared with controls, we have already seen discrepancies which may well be due to differences in processing of cells (ie cells isolated and tested fairly quickly vs cells cultured for several days before testing). This is discussed
here and
here and at various points in these
two threads.