For systems using radioactive decay as seed, this should have very little effect, as the involved previously-unknown effects are extremely small and the properties influenced by them are unlikely to be used as a source of randomness anyway.
It's like if I said:
"What if zebras are green and red, but we see them as black and white because we don't drink enough water?".
It kind of sounds like an actual logical argument, but it's not even wrong.
Why did you jump to the conclusion that the well understood thing should be replaced by the new and unknown thing?
Perhaps you mean that radioactive decay may have a causal mechanism involving dark matter? Possibly. For such a model to be taken seriously, it would need to agree with our (presumably very numerous and precise) existing observations, as well as either predict the outcome of some new, falsifiable experiment; and/or simplify/unify some disparate existing results into a common framework.
Even if such a model manages to overcome all of those filters, radioactive decay would still be a very real phenomenon. For example, discovering that light is an electromagnetic wave doesn't at all imply that "there's no such thing as light"; discovering that gravity is curvature of spacetime doesn't mean "there's no such thing as gravity"; etc. Like radioactive decay, those are phenomena which we have very direct evidence for, and which theories and models must try to explain.
In comparison, some theories/models postulate the existence of entities which we have no direct evidence for, such as phlogiston, luminiferous aether, dark matter, etc. The only evidence for these is the success of the model that predicts them; that evidence may be undermined if a better model is discovered which doesn't require these entities (e.g. the kinetic theory of gases for phlogiston and special relativity for luminiferous aether).
Hence, if we discover a better model for predicting flat galactic rotation curves, gravitational lensing in excess of the visible matter, etc. which doesn't require an exotic form of matter, then we might claim "there's no such thing as dark matter".
However, even if that were to happen, we certainly couldn't claim that "there's no such thing as flat galactic rotation curves", "there's no such thing as gravitational lensing in excess of the visible matter", etc. since, like radioactive decay, those are phenomena we have observed.
If a theory disagrees with observed phenomena, or claims "there's no such thing" as the observed phenomena, then the theory is wrong. That's precisely what empirical falsification is about. We've observed the phenomena, so that theory cannot describe the world we live in.