Here's your only real mistake. It's pretty serious.
Flats ADU should be _much_ larger than lights ADU. For example, take the D5300 (yours) and the D5500 (mine, identical for this purpose).
Lights ADU should be in the vicinity of 1000 ADU (16 bits), raw. Bias subtracted, 400 ADU (16bits). ISO 200. Getting it exactly at 1000 has no value whatsoever. The linear histogram is a thin spike on the far left.
Flats should be about 32000 ADU (16 bits, which has 64000 as the maximum possible ADU). The linear histogram should be about in the middle of the graph. Getting it exactly at 32000 has no value.
If you shot flats the same ADU as your lights, it's no wonder your corners were too dark. I'd expect them to be.
Needless to say, the calibration software makes this all work. BUT, you must include either bias or dark flats. Or it won't work.
Yeah that really was a huge mistake looking back knowing all I do now.
When I opened a really overexposed flat up in Iris it looked like the maximum for a pure white linear frame was ADU: 16383 - Density: 24160256
Everything else was zero..
I only downloaded Iris several days ago when started learning about all of this, so I am definitely still learning the software. Would 16383 ADU being the max signify that I am looking at my photo in a linear 14bit depth?
After I tuned the exposure back a bit I got the peak of the histogram on Iris to sit at ~8000 ADU, I figured this was pretty perfect since ~16000 ADU seemed to be the max.
I always use bias, but I did not think about making the firm point that bias or dark frames are absolutely necessary when taking exposures like this for everyone else reading.
I am happy to hear that I have a pretty good foundation of the concepts though! I really can't wait to see how big of a difference this will make for me on my next trip out to my local dark site.
Did anything change after modification of your D5500? I know you will get to 1000ADU quicker on your light frames after modification but was there anything else that changed?