All this talk of zillions of watts as a bad thing really comes from not understanding thru experience that all those watts don't translate into crazy speeds. And further, how electricity works insofar as motor wattage is concerned. For example:
Wattage output is a function of volts x amps. So even a lowly 36v battery, which charges to 42v when it is at 100%, mated to a very tame 20a controller (very commonly in use on manufactured ebikes; mine came with that combo in 2017). 42v x 20a = 840 watts. Not 750w and certainly not even 250w. Furthermore, that bike had a motor stamped 350w on the outside. Once you understand how this works, it should very quickly become obvious to you that almost nothing in the EU conforms to the 250w limit. In fact, manufacturers largely stopped advertising 250w output and instead started substituting Newton-Meter output in their advertising. Thats an unregulated number (absent the 4x output rule that isn't quite EU law IIRC).
Using the math above, go ahead and try and find a way that a bike with a 48v motor can possibly conform to 250w output, and more importantly read the EU laws (still used by the UK) that govern what allowable peak output is; never mind the average output measure. A 48v battery is 54.6v charged to 100%, and how do we keep that peak output under 250w? Well, a measly 5 amp controller exceeds the limit at 273w. So even if you do some convoluted gymnastics to come up with an average output rating, the 250w limit is exceeded by everyone. Period. The same is largely true of 750w in the USA and there is ZERO regulatory appetite to change this, since an ebike with 'legal' power is gutless, nobody would buy them, and ebike adoption is a priority for the government at pretty much all of its various levels beyond the local city council. Crushing the life out of the platform is not going to happen.
The toothpaste is already out of the tube on this subject; especially here in the USA, so get used to it.
About the not-understanding of power and how it translates to actual speed, by way of simple example, my full-power twin hub commuter, which has two '750w' motors powered by two independent controllers, each rated for 35a and both powered by a single 52v (58.8v on a 100% charge) battery, lets see how that translates on paper: 58.8v x 35a = 2058w. 2058 x 2 = 4116 watts. Also since both motors are rated for 80 Nm thats 160 Nm of torque. But of course these are peak numbers and not continuous, first and foremost. Next, I have dialed the controllers on both motors to use full amps but slow-start so power rolls on sanely, and...
... since I am a pedaler and not a throttler, the two motors are both tied to one pedal-assist sensor, and my pedaling puts the bike at a top speed of about... 28 mph, which is legal for an ebike in a bike lane here. Furthermore, I have geared the bike so I have a very large chainring in front and a very small one in back. So I can pedal up and over the 28 mph limit to about 34; especially on the way home with our normal afternoon tailwind. Nobody cares, and that includes the traffic-enforcement officers stationed roadside who have radar'd me, and more than a few squad cars that have paced me on a 35 mph road where they can do it without being TOO obvious. I'm pedaling, wearing a helmet, I have head and taillights. I obey traffic rules (stop lights and so on).
So, 4000w is not some bugaboo Millenium Falcon. Doesn't work that way. BTW my hill climber cargo bike, also 2wd, has 20a out front, 30a in back, 52v battery and the math on that one is 2940w. Its top speed is 25 mph if I really REALLY work at it, but since it goes up hills, it spends a lot more time at 10 mph or less (on a pedestrian path). Bicycle gearing matters.