[__MatrixMan__]

I've always wondered how quickly this effect would ruin the monoculture.

Suppose you're treating the engineered microbes like sourdough starter: Add them to nutrient solution, wait for desired effect, take a small sample, add to next batch.

You'd expect a selective pressure away from the engineered outcome, such that after X batches, your samples are overrun by wild-ish type mutants. But is that a merely academic concern, something that might happen after 500 batches, or are you out of luck by batch three?