dsws

Maybe it's not always a matter of finding the proper signal for germination. Maybe sometimes it's a matter of enabling marginally-viable seed to germinate.

That seems as though it could be harder. You would need to understand not only the signal but the recognition and growth processes, so as to support the step that fails. Or it could be easier: maybe one aspect fails most readily, across a wide range of conditions, so that you don't need to know anything else, just what it takes to support the weak link.