I realized I can get a good estimate of the total output (lumens) from my measurements with a bit of math trickery. Because the radiation from the LED is distributed in angle in a known pattern (it is very close to a Lambertian surface), I can derive the total output if I know the luminous intensity directly above the LED. I did this calculation when I was trying to get an absolute value of the output from a ceiling bounce test.
I will outline it here:
Just multiply the luminous intensity (in cd) above the LED by pi to get the total output in lumens.
As a sanity check I will compare the output of the XPL V6 1A using this method with the XPL V6 2C that djozz measured, at 4.0A. LEDs with dome are also very close to Lambertian surfaces so this should be a good estimate. From here, the luminous intensity above the XPL V6 1A with dome at 4A was 522.7cd. Multiplied by pi we get 1642 lumens, which is close to the ~1575 lumens that djozz measured at 4A.
I added the plot with total output for the XHP35 to the OP.