Reproducibility is indeed important. From the point of view of those in
community and in particular in the R community, the key issue is that
the data on which
a publication is based should be readily accessible so that others can
possibly extend the analysis, and propose and present alternative analyses.
But to say that "most published scientific research is probably false"
nonsense. That sort of assertion is made by right-wing ideologues who
with envy of the academic community and slang off at it to alleviate
their own sense