That would certainly lead to people checking their references better. But a lot of references are just in passing, and don't materially affect the paper citing it.

One would hope that if some work really did materially depend on a bogus paper, then they would discover the error sooner rather than later.