I was meaning a collapse in the population of the algae. When using algae in an industrial process, they would want it to be continuously feeding and reproducing, at as high a rate as possible. This means that it must live in an environment that is highly tuned regarding water quality, temperature, and its food inputs. Any problems in these parameters and the algae will die out fast, not gradually. The time between a problem occuring and a collapse of the population could be as short as hours, not days or weeks.
If algae otherwise could be coaxed into yielding passable efficiency to a fuel process, the growing sensitivity of the algae probably would be the least of our worries.
I don’t think this could really take off until we came up with an algae that thrived in the saltiness of ocean water. The expense of furnishing it with fresh water would be prohibitive.