The GLM mic likely requires plug-in power which your audio interface cannot supply.
You could either buy a phantom power to plug-in power adapter, like the Rode VXLR+, or you could use something like the Apple headphone adapter as your interface, as it is designed specifically for these kinds of microphones and supplies plug-in power out of the box.
Furthermore, GLM likely loads a calibration file for the mic automatically, which you likely don't have direct access to.
That means that measurements done in REW without that calibration file would be less accurate.
As a fix, you could setup the GLM mic on a sturdy stand, then do one measurement in GLM and right after, one in REW.
Ideally, once you've plugged everything in, you would leave the room, close the door, and start the measurement remotely using for example Chrome Remote Desktop, because you standing/sitting in the room could impact the measurements and introduce additional variance.
After you've done measurements in both GLM and REW, you can export or just screenshot the GLM graph (try to get a version with minimal smoothing applied), then convert it into .csv using
https://usyless.uk/trace/ or
https://automeris.io/
The .csv file your can then drag and drop into REW, at which point you'd try to smooth your REW measurement as close as possible to the GLM graph.
Finally, Trace Arithmetic -> A/B in REW would output the approximate mic calibration response that GLM applies in the background.
To double-check, you can export it as .txt, load it back into REW as mic cal file for your audio interface in the preferences, then perform another measurement (again ideally w/o you in the room), and that should now match your GLM graph.
This is still not perfect as we don't know what kind of windowing GLM applies to its sweep recordings, so cannot set REW to match, but it's the best I can come up with.
