Thanks for your comments/suggestions Mads/Jared,
Since it was raining today, I tried a few different settings. The best that seemed to work well was:
High Dynamic Range (Gain 0; Offset 50); first brightness setting on Spike-a-Fielder
5 secs for Lum (1x1); RGB and NBP (2x2). Based on the expertise of others, I've started with 2x2 binning of RGB and then resize then in Nebulosity before layering in PS.
For the 2x2, I could go as low as 4 sec without getting horizontal banding, except for Lum, which needed 5 seconds minimum.
If I increased the brightness of the light box past the first setting, the Lum was totally blown out. Similarly, if I use the Unity setting, the exposure time would be below 1 sec with heavy horizontal banding and hence sticking with the pre-set HDR setting.
I tried the Flats Calibration wizard and the only settings that seemed to work (all green calibration check marks) were Target ADU (31,000 +/- 500); 0.5 sec minimum; 10 sec maximum. If I used an ADU higher or lower than 31,000, one or more of the filters would fail to calibrate.
The problem after saving the calibration is that setting up my sequence for flats, SGPro would put a calibration factor in for the Exposure and all of the exposures were set below 5 seconds and significant horizontal banding occurred.
I then ditched the calibration wizard and ran the Frame and Focus until I got what appeared to be good flats without banding. (see below). I like Jared's suggestion of rotating the camera and see if the vignetting stays in the same place. I was wondering if it had something to do with the back focus? I use a 0.7x reducer on my Celestron EDGE HD 800; a T-adapter, FW and ZWO camera. Measuring the distance, Im within 1-2 mm of the Celestron's reference 105 mm back focus from the end of the 0.7x reducer.
Flat_5 sec_1x1_Lum_frame 10
Flat_5 sec_2x2_Red_frame 10
Flat_5 sec_2x2_Green_frame 10
Flat_5 sec_2x2_Blue_frame 10
Flat_5 sec_2x2_NBP_frame 10
David