We seem to be reaching good enough models (arguably some of them will be thoroughly reproducible in laboratories soon enough). But I don't think we'll arrive at a definite answer here. My guess is that we'll come up with multiple ways it could've happened, along with their respective degrees of likelihood.

/me leans back on his armchair