Very likely. Normally we can assume that if the base station is not where we think it is the base station didn't move and so therefore that is how far off the current GPS signals are and so we can tell the tractor to correct by the same amount. However if the tractor and base station are receiving different answers we have no way to detect that.

> GPS accuracy issues are solved in trillions of devices around the world.

They are IGNORED in trillions of devices around the world, not solved. Normal GPS is at best to within a few meters. Most of the time that is good enough - if you are within 10 meters of where you want to be most people can look up and see the door and get there. (the blind would have trouble). If you are trying to guide something to within 2 cm GPS is only solved if you can make some assumptions that normally hold by didn't. This isn't a bug - at least not until someone can explain how we can fix it.

I work for John Deere, but do not speak for the company. I do not work in the GPS area, I've talked to a few of those engineers but I don't have any more knowledge that is public.

It is indeed a bug, for John Deere has sold millions of farmers gear dependent upon cm positioning, each farmer spending enormous amounts pf capital for this gear, made enormous promises predicated upon this, and so clearly it isn't impossible.

Right?

(your best option here is to not answer, I assure you)