The amount of testing (and testing with accuracy) would be extremely high. Even if it's possible, I can't imagine the cost would be remotely feasible given the sheer number of ways to track someone without searching for their poop in sewers.
For example looking for populations that aren't leaving houses.
It's a far-fetched scenario, but I don't know enough about the breakdown of DNA in wastewater to know if it'd be possible to locate him via this method. If he wasn't shedding a unique strain of a virus, let's say, but where his genetics are known.
This researcher explained to a reporter how they conducted a binary search of sewer lines from the treatment plant, all the way back to a single manhole which was only fed by a single employer with 30 employees. I just don't know if this is only possible using viral RNA, or if anyone's DNA would be possible to search on in a similar manner. For good public health reasons, we're likely to have continual wastewater surveillance at all major plants going forward - I just don't know how far it could be pushed.