How do they measure the U-value of windows?

How do they measure the U-value of windows?
I've been trying to correlate heat flow measurements using double pane windows as the testbed. The windows are rated as U.30 >> R 3.33.
The manual for the Omega OS650 says to measure the temperature of an interior wall and a window. Heat flow thru the window is delta-T * BTU/hr/ft^2.
The heat flow thru the window should be the difference in glass indoor/outdoor surface temperatures * U. Those heat flow numbers are not anywhere near the same. I added R-5 foam insulation on the outside of the window and looked at the ratios between R3.33 and R8.33. Still no correlation.
The glass surface temperatures are different from ambient on both sides. And that's critically dependent on convection, wind etc.
The first piece to the puzzle is, "How do the window people come by that U=.3 number?
NFRC100 claims to describe the process, but all it says is, "use the computer model. Measure it only if you can't." But falls short of describing the measurement technique. Maybe it's in the references I can't access.
Other references suggest that the boundary layers on both sides of the glass are significant, but the numbers are all over the map and unclear (to me) what they're describing.
So, when the window vendors measure the window, what do they use for delta-t? Glass temps? Ambient temps? Something else? Using what assumptions for airflow?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
Add image file
Upload is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.