The above video is a rather ridiculous way to get hydropower.
He has a non-stop flowing creek, and he diverts a small bit of it through a hose, to get high pressure, and then only yields 25W? Meanwhile all that potential and kinetic energy from the water is spraying about. He is literally reinventing the wheel, and ending up with a poorer version of it.
Even if you want to keep that design, you want to do something better to prevent all the splashing, as that's energy that could be used. Ideally, all the water dumps out at the bottom with very low speed (having shoved all its energy into the wheel). Expecting to turn a motor at its operating speed is silly, as at those speeds, it's better at throwing water from centrifugal force than catching it (as you can see in the video).
A better design, and one that would yield (guestimating) 50-100x as much power, with smaller negative impact (no diversion), would be to place a much larger (several feet in diameter) paddlewheel (with sides) into the stream itself right below the reservoir (where the water is already falling). Have 100% of the creek flow fall into the top of the wheel and then get dumped out the bottom and continue on its way. He would have to do some rough calculations for flow rate of the creek (a barrel and a stopwatch). Then he could build a suitably sized wheel, and using the flow rate of the water and diameter of the wheel, pick a suitably-sized motor. Perhaps a truck alternator would be a good choice.
The paddlewheel of course would spin slowly, but that is okay. Using pulleys (easy to make out of wood, they'll be big) and belts, or a chain & sprocket, or gears, gear the paddlewheel up to the appropriate speed for the motor. If properly designed, it won't (can't) stall out from lack of torque because even if the flow rate drops, no matter what the generator can still be turned over by the weight of the water (so maybe each section of the wheel fills up slowly before it has enough mass to turn the generator over a bit).
Also, this presumes zero water speed, you're designing purely on a known mass of water dropping the height of the water wheel per second (or half the height I guess, until the water starts to spill out). So if you get any water speed coming into it, that too will be extra beyond the minimum you've designed it to be able to turn over at.
You could design it differently too, you could have a stream-dipping waterwheel (wheel above the stream, using only stream speed and no water weight), lots of options, depending on the situation.
Farmers have been using these systems to generator electricity (or mechanical power, for a self-powered irrigation system) for decades, and watermills have been around for centuries doing the same thing, it's just a matter of getting electricity out of the shaft.
It's far easier to ship electricity (wires) than it is water (pipes) too. And wires don't freeze or need to be filtered or get clogged.
Also, aesthetically, most people would probably find the sight and sound of a gently rotating wheel by a cottage more fitting, than something that sounds like a cross between a motorboat and a wet fart, and looks like a headbanging punkrocker. But that's entirely subjective.