Small diversion into the world of academic research... CALTECH have had a group looking at improving MEMS devices for inertial sensing. They developed an idea where the position of the reference mass in an accelerometer was measured and cooled in a laser trap. The basic idea was to remove as much thermal noise from the device as possible. From memory they'd got it good enough to give very good drift rates, something like a few meters in a half hour. [An application for it discussed in the paper I read was that your mobile phone would still be able to work out which aisle you were stood next to in a supermarket half an hour after you'd gone into the shop]. Their plan then was to go onto the same idea for MEMS gyros.
Here's an article about the work of that group:
https://www.electronicproducts.com/S...lerometer.aspx
And I've heard nothing substantial since. Either it didn't pan out, or it couldn't be productionised, or someone governmental got wind of it and worked out the consequences of mobile phones having INS that good. I've not been keeping an especial eye out, but I'd expect to have seen stuff in the wider tech press by now had actual devices been prototyped.
I think laser fibre gyros are generally not quite as good as ring cavity laser gyros due to the dispersion of a laser pulse in the fibre. I may be wrong, but I think there is a compromise between the number of turns in the fibre coil (more is better for sensitivity) and the dispersion (which is proportional to the length of the fibre). Still, they're pretty good as these things go.