in a (made up) word: TANSTAAFL (there aint no such thing as a free lunch. thank you heinlein or nivin, I forget which)
take a step back and look at the system engineering approach.
while what you propose COULD be done, it would be a net energy LOSS, not a gain.
(except in certain very small cases, like individual homesteads where the conditions are just right; and there, the intermittant and very small (on an energy production scale) flow rates don't justify the expense of the installation.
as an example, calculate the amount of energy available in any reasonable water pressure (say 4 atmospheres, (equivalent to a head of ~120 ft or ~40 m) which is a high average municipal pressure) at a flow rate of (say) 100 gallons (~400 liters, or .4 metric ton) per day. the proof is left as an exercise for the student. <G> (hint- the answer is not in kWh/D, even if there is no residual pressure for the point of use- no showers and no water on the second floor at all))
(100gpd is higher than MY total water consumption, and my biggest vice is longish showers. . .)
as for "why not generate power from 'domestic water' supplies:" in most urbanized parts of the world, water is supplied from wells (requiring much energy just to pump the water to the surface), rivers (requiring pumps just to create flow and to feed any sanitation plant) or from distant reservoirs (requiring much engineering to get the water to where it's needed with minimal pumping- as an example, san francisco's water travels something more than 100 miles from point of collection to point of use)
in any event, the amount of energy required to get the water to point of use has been optimised to its lowest value consistant with reliable supply.
as a matter of fact, the most of the earliest portions of the industrial revolution was tha application of fossil energy to the movement of water and sewage. (pumping water out of coal mines . . . . followed by pumping river water for domestic use and then pumping sewage back into the river for disposal; the energy contained in urban water systems was put there through the planned action of mankind. THERE IS NO EXTRA, or the design engineer/s will be looking for a new job. (as a Walmart greeter, perhaps?)
just to drive home the point, and to paraphrase the laws of thermodynamics:
you cannot win (gain energy over inital conditions)
you must lose (no piece of equiment is 100 percent efficient,)
and you cannot get out of the game (all energy conversions suffer from entropy generation; i.e. all conversions are a net loss over and above the machine loss.)
so, the energy you wish to capture must be put into the system at some point. a positive pressure is maintained so that any leaks do not provide an entry point for pathogens; that positive pressure requires additional power input.
flow losses eat some (most) of the input power. (come to that, leaks do too)
and running a turbine on the residual loses some to entropy, some to efficiency, and lowers the pressure you need to get your toilet to fill, etc.
the upshot: there simply isn't any addtional energy to be had in a large municipal system, and further, an electric power line is a lower-loss tranmission system than any large-scale hydraulic system can hope to be.
however, it may interest you to know that a freind of mine, who builds and maintains the controls for a large municipal sewage treatment plant, reports that said plant is totally power self-sufficient by the simple expedient of burning the methane produced by the sewage in large piston engines. (think of the size engine that powers a container ship) they even manage to sell a small amount of power back to the grid to offset the cost of running the plant. (waste-water treatment is a growth business, gang. . . more toilets are connected every day! he considers it to have been a short week if he works less than 90 hours)
-Dan