> No matter how low the odds are, the counts of those potential interactions bring this outcome to a certainty.

That is simply not true. Events are classified by probabilities, and there are a whole lot of things with a probability a lot less than will happen across the whole ocean across however many billion years.

All the higher-probability events will occur, yes. But a specific sequence of multiple extremely low-probability events? That then continues to replicate before it gets wiped out by chance?

Not a certainty, absolutely not. Contrary to what you say, it matters very much exactly how low the odds are.