The problem with (or the advantage of) the water flowing analogy, or even more broadly the discrete element model, is that it explains reality good enough to be used in most practical situations. Schematics are ubiquitous, yes they are "fake", but they are also usually "correct enough". Kind of like the incorrect Bohr's model of electrons orbiting the nucleus actually does explain the emission spectra (up to a point).
There is one commonly used concept that requires understanding electricity correctly, and not just as a combination of waterhoses and gizmos. It's impedance, and it directly corresponds to the "controversial" experiment that Veritasium is proposing in his video. Impedance breaks the pipe-of-electrons analogy.
Yes you can keep going a bit further, but that's still the lumped element model. The problem is when you analyze something like a transmission line - like that circuit presented in the video. Or a PCB with very fast signals where you have to understand that the energy moves through the insulator, not through the conductor, or the circuit will not work.
One way to look at this is that there is no such thing as a hose for electricity. It cannot be confined to a conductor, even if it is also wrapped by an insulator. It is only mostly confined. And this is not some failure of materials engineering that we may overcome one day, this is just how this stuff works.
BTW, the answer in the video is 1/c seconds, i.e. one meter worth of speed of light. And the lightbulb will experience current determined by the impedance of the transmission line. Then the fields will do a full wraparound the ends, at which point the circuit will start stabilizing around the resistance of the load. It can take a few back-and-forth iterations to stabilize the current.
I have gone down this hole many times before and while it is kind of possible (the equations for a capacitor and an inductor are basically just a spring and a flywheel), it just creates really convoluted images that won't fit well (or will be too convoluted) when you try to integrate into wider electronics.
Yes, all pipes leak, even DC pipes, in two almost entirely independent ways (electric and magnetic). And the vector product of these leakages (the Poynting vector) is what actually transmits power. Note that this energy transfer happens entirely outside of the conductor.
Sounds bizarre, right? That's why this is mostly ignored unless it can't be, for example in very fast circuits.
But there is an accessible video that explains electricity pretty well. Veritasium - The Big Misconception About Electricity: https://www.youtube.com/watch?v=bHIhgxav9LY
There is one commonly used concept that requires understanding electricity correctly, and not just as a combination of waterhoses and gizmos. It's impedance, and it directly corresponds to the "controversial" experiment that Veritasium is proposing in his video. Impedance breaks the pipe-of-electrons analogy.