What has helped the most however was adding in a humidifier. Dry desert air is unforgiving and adding in water to the air made a huge difference in comfort. 68* F with 40% humidity is much more comfortable than the same with 5%.
The data recording setup they are typically using only works for air-to-water heat pumps, like you might use to heat water for radiators. They rely on a 'heat meter' that measures the water flow rate and temperature.
Typical setup: https://shop.openenergymonitor.com/level-3-heat-pump-monitor...
If your system is a typical US air-to-air 'split' system, you probably have a long loop of refrigerant in a pressurized tube to carry the heat. To measure efficiency, I think you either need to measure air flow accurately inside or outside, or measure refrigerant flow rate. Either one harder than the commercial meter for hot water.
So, no easy drop-in COP measurement for me.
And the same electricity can raise the temperature by even more degrees Fahrenheit!
What we really need is a combination of the two. Something that measures air temperature and water content because 68F at 5% humidity is a lot different than the same temp at 40%>
Nothing beats scientific accuracy and thoroughness, right? So it then actually ended up being tied to water as well:
> For much of the 20th century, the Fahrenheit scale was defined by two fixed points with a 180 °F separation: the temperature at which pure water freezes was defined as 32 °F and the boiling point of water was defined to be 212 °F
The "feels like" apparent temperature accounts for things like humidity and windchill[1].
Many weather apps provide the "feels like" temp, including my app: https://uw.leftium.com
I was going to drop the "feels like" reading in my new weather app (I just didn't notice a major difference), but maybe I'll keep it...
For some reason I never noticed there are exactly 180 degrees between freezing and boiling points on the Fahrenheit scale. 100°C is a nice "round" number, and 180°F divides evenly into a lot of smaller numbers.