@OC: I think your "small percentage (of 7200) able to read up and find those automated methods of aborting work" after they've done the points pay rate calculations would not be enough to explain the numbers of aborted FA@H WUs that I'm seeing. Remember, my 5 machines are only a tiny percentage of the many thousands of fast returners crunching the project.

To explain to readers what OC means when he says "more resources are used when re-sends are invoked and that is a waste", I noticed in December 2013 that if I abort a zero-redundancy FA@H AutoDock WU, 2 repair copies are sent out, even if I had received the WU only moments before aborting it. The 2nd repair copy is then a waste of the crunching time of the device that gets it.
(I asked about this in thread FAAH Vs FAHV Replication )
Thus the people who are aborting FA@H AutoDock WUs are causing lots of wasted crunching for the rest of us. However as OC points out, and providing that a large percentage of the group doing the UAs (who may not be Ripplers, Ripoffs or whatever) is not pulling out of WCG completely, there is probably a net gain to WCG. [Edit] Just checked WCG Supplemental Performance Charts - 1 >> CPU Years by Day & Research (Last 3 months) . It shows that since Nov 13, WCG daily totals have increased from around 400 CPU yrs/d to about 750. The increase in work has gone to MCM. Someone big has arrived, anyway! [/Edit].
If they crunch only projects other than FA@H, eg MCM, those of us who have both FA@H and MCM selected will get fewer MCMs and more FA@Hs, meaning more FA@H work will get done so they'll be contributing indirectly to FA@H anyway :) . (Shh-h-h-h!)

It would be logical that if aborted zero-redundancy FA@H AutoDock WUs result in 2 repair copies, then this should also apply to FAHV WUs as well. Looking at my Results pages, it seems that this may not be the case: I found examples where my copy validated when all other wingmen aborted their copies. (Anyone else confirm?)