Having worked with many graphic film developing machines which ran a double shift 5 days a week. I have heard and seen many conflicting ways to filter water.
Pretty much the best situation is to monitor the usage or throughput, of water actually passing through the filter.
Our DuPont machines, had a meter to measure the amount of water. When a certain point had been reached, we changed the filter.
I myself have a 5 micron filter on the input side of the darkroom. Basically it's there to filter out particles large enough to make a difference to ensure film doesn't get particles stuck on it's surfaces.
I change the filter about 3 times a year in a busy year, twice in a slack year. When I find that there is slime, I figure that the filter has been in too long. Not that, that is a correct scientific situation, it's just what I do.
I've been doing this for the last fifteen years.
Mick.