John: If you don't trust radioactive decay, I suggest chunking your atomic clocks. (Or rather, the ones synched to one). They're useless junk to you.
Hanani: The only thing that can contaminate a radioactive decay series (to lead to a wrong date) is it's
daughter element.
Luckily, that's not a problem. There are at least
three ways to be sure a sample wasn't contaminated.
1) Date by multiple methods. After all, the odds against multiple daughter elements being present in
just the right amounts to force all the clocks to the same wrong date are pretty inconcievable.

The odds of it happening across multiple samples, well....no real worries there, eh?
2) Use isochron dating. Isochron dating is a nifty form of dating that well easily show if a daughter element was present.
3) Use a method on a sample where you can rule out presence of the daughter element. For instance, some rocks are formed in such a way that no daughter element could be present. No worries there, huh?