Hi @moeythomas . We agree that having to fudge your elevation to calibrate your sea level pressure is NOT THE RIGHT APPROACH…and to be clear, this is not the approach that WF is taking.
Here is how it actually works: AIR will sense the pressure at it’s location (station pressure). To normalize and compare against other stations in the area, total elevation needs to be taken into consideration…which is then used to calculate sea-level pressure…so that everyone is comparing apples to apples. For your WF system, total elevation = [elevation at your location automatically determined when you set your geo-location] + [height above ground at which the AIR is mounted].
In-situ sensor calibration is typically done by comparing known good data sources and applying a calibration. Unlike Davis et. al. we don’t think this should be something that the average home user (or anyone for that matter) should need to do. INSTEAD, the WF auto-calibration / continuous learning (CL) system will automatically identify and monitor a variety of known good data sources for your immediate area and automatically compare them to your AIR data on an ongoing basis and apply any required calibration.
As mentioned previously, the WF CL system has not yet been fully deployed. As of today 19 Sept, only RH is run through the daily CL calibration system. Other parameters including Pressure is in final testing and will be rolled out soon.
So there you have it…we agree! Have faith, we’re not idiots. 
Will announce when the CL system is deployed for Pressure.
And PS - we suspect that a batch of AIRs did not receive a proper pressure sensor calibration at the factory which might explain the 2-3mb deltas being reported.