Some ideas/tips I had a long the way.
I wish there was a bit more extensive self test during startup. Most of the issues I had were the belts going slack after cutting for a while. It would be nice if it could move around and test the belt tension to verify the calibration is good.
I would love to have a start job on line function. Maybe this exists but I could not find it.
The button pusher from the not store is a life saver! Basically required for any tool change. I also should have gotten one of those remove before flight tags. I left that in a few times and made the router unhappy.
It’s very easy to over tighten the bolts in the circle part of the arm causing increased friction. I must have missed this in the build instructions but this caused us a lot of issues early on. Also greasing the large gear that holds the belt was pretty much required on all of the arms we had. I would love to see these added to a troubleshooting guide.
The best way I have found to take it apart is to lower the z on M4 until you can get a allen wrench into the ‘Support Riser Top’. Undo all the screws on the top and then send the z all the way up until it comes loose. It saved us a lot of time and it’s way easier to access these screws.
Dust is a killer. There was a couple time where our dust extractor failed. I am pretty sure this dust got into the z motors and then they stopped going up. Then the router pretty much bottomed out and carved a nice pattern in my concrete floor. hehehe. To combat this I noticed dust would be pushed out when the motors went up. So I increased the z clearance while traveling to 12mm hoping this would flush them out. I noticed a bit of an improvement but I am not sure how conclusive the results are.
I want to design a 3d printed part to cover / clean the screws before they push stuff into the stepper motors. I also noticed after they got dirty they would get very hot. So I also slapped some heat syncs on them.
I would also love to get a z encoder to stop the thing before it bottoms out. I am not totally sure how you would design one that could deal with all the dust though.
The persistent problem we had was the belts going slack after a few cuts. I think it would take about 9 hours of cutting or so. After that we would have to recalibrate to keep the belts taught.
One thing I have been thinking about is that my spoil board is 3/4 particle board. It was mostly flat but it does curl up a bit on one end. I didn’t think it would be a problem since it flattens out when the maslow moves to that area and pushes it to the floor.
Oh also none of my circles were round 
it didn’t affect my project but I am going to have to do some measurements to figure out the right correction factor.
Well this is getting long… Bar if you read this I would love to get coffee with you can chat, I am Seattle based. I would love to contribute some of these ideas to a troubleshooting guide but I am not sure where to start.