It's funny how this cost benefit calculation is considered valid when applied to research on mice, but any such calculation used to try to justify research on humans would give a "divide by 0" error. I think there's an inconsistency there somewhere.

"it can't be done any other way" strikes me as pretty thin. Whereas we may not be able to conceive of another way at this point in time, it does no good to imagine that this condition will continue in perpetuity.