As documented in Fixing CSAT3s, Steve found a method of restoring data from the CSAT3 sonics. I tried to figure out if we could get rid of the warning message about mismatched signatures, in case that was related to cal data or something else that was still corrupted. That is documented in a JIRA issue: ISFS-140. The end result is that Steve looked at data on a DSM for four CSAT3 sonics, and now we're convinced that they work:
1119, 1120, 1122, 0538
1121 was never connected to a DSM with oversampling enabled, so it should still be fine, and will be tested when staged with a tower.