That's one aspect of the problem yes . That and the lies. A lot of chemistry papers are simply not reproducible.
But that's different in industry, where only few companies share things.
I wish there would be some kind of data broker for companies that would release data publicly when companies die or abandon projects.