PMs work with starin gauges and bathroom scale work with strain gauges.
I have a bathroom scale the gets down to .01 lbs and it cost me $70. I have a PM that measures down to the .01 ft/lbs and cost $4000. One was manufactured in Germany and one was manufactured in Asia. Granted, the PM does velocity as well but so does a $40 cycling computer.
You have a bathroom scale that does 0.01lbs? Mine does tenths.
Another thing to consider about PM is that you need 0.01 accuracy over a much higher range. No bathroom scale I know of reads 2000lbs with that kind of accuracy. I'm no expert, but I think that proper strain gauges like the ones used in an SRM are the biggest cost. I don't pretend to know this for sure. I'd like to hear an answer to that from someone more in the know.
The other factors such as reliability, spinning gauges, waterproofing, and battery life are also added in there.
A number on a digital display means nothing at all. A lot of people think more numbers is more accurate, but it isn't so simply. Let's take a SRM as example and a powerfull sprinter. The accuracy of the SRM is -/+ 2%, he is sprinting at 2000 Watt. So the readings on the screen can be somewhere between 1960 and 2040 Watts, quite some difference. But on the screen you can see changes of 1 Watt, when that 1 Watt would have been significant the accuracy would have been 0,1% or even better.
The 8 strain gauges in a SRM cost probably less then €10. SRM's are just way over-priced. €900 for the computer, which isn't high end at all... Programming the microcontrollers isn't rocket science, PCB layout is also not too hard for such a simple device. You don't need to have a masters degree in EE to develop a powermeter.
But there are more peeps working on a affordable powermeter.