I moved it here from eBikes. Sorry, we usually leave a note explaining

Not since May 2000, AFAIK. Look up "GPS Selective Availability."

"Initially, the highest quality signal was reserved for military use, and the signal available for civilian use was intentionally degraded (Selective Availability). This changed with President Bill Clinton signing a policy directive in 1996 to turn off Selective Availability in May 2000 to provide the same precision to civilians that was afforded to the military. The directive was proposed by the U.S. Secretary of Defense, William Perry, because of the widespread growth of differential GPS services to improve civilian accuracy and eliminate the U.S. military advantage. Moreover, the U.S. military was actively developing technologies to deny GPS service to potential adversaries on a regional basis.[SUP][26][/SUP]
Since its deployment, the U.S. has implemented several improvements to the GPS service including new signals for civil use and increased accuracy and integrity for all users, all the while maintaining compatibility with existing GPS equipment. Modernization of the satellite system has been an ongoing initiative by the U.S. Department of Defense through a series of satellite acquisitions to meet the growing needs of the military, civilians, and the commercial market.
As of early 2015, high-quality, FAA grade, Standard Positioning Service (SPS) GPS receivers provide horizontal accuracy of better than 3.5 meters,[SUP][27][/SUP] although many factors such as receiver quality and atmospheric issues can affect this accuracy."

Here is a typical ride, Elev Gain and Loss are the same.



For the ride last week (that started this thread) it was off by ~35 ft, not huge, but not normal.

The issue was that I posted to e-bikes, this thread was moved almost right away to electronics, which is where I intended to start the thread.

I suspect the gain/loss measurements are more accurate for larger values. The noise in the readings might be close to real readings on a flattish ride.

800 feet is on the flat side.

In the case of your first problem, if the two rides were basically the same (they aren't the same), it appears the barometer became less responsive during the ride (the profile becomes progressively flatter). It appears it was fairly unresponsive at the start (the scale might make that hard to be sure of).

Ah, I see what happened.