I think the biggest question here is to what extent can productivity in basic research be quantified. Or, what are the value to failed experiments and explorati

I think the biggest question here is to what extent can productivity in basic research be quantified. Or, what are the value to failed experiments and explorations of "out there" ideas? Basic research is tricky because it is at the periphery of human knowledge, so getting our bearings there is difficult. The focus should, you are right, be on "overall productivity" and not on individual productivity, but the question is how do we make decisions about individuals based on overall productivity goals.