I mean that the gas model didn't crumble the first time the predictions didn't match the evidence. The test were changed to test for other options and other possibilities. Whether these specific changes applied accurately to this question and whether such changes eliminate other models remains to be seen. That's what I mean.