I wasn't saying everything therefore lights up, just saying that a 50 degree difference between night and day, doesn't mean a 50 degree difference in all surroundings heat signatures.
The distribution of grey scales is probably dynamic, so you can then have as much contrast for day as night. Upper and lower bounds may also be dynamic, so let's say you have from -30 to 90 degrees at night, 30 to 90 at day. A certain number of shades of grey on the greyscale, let's say 60 for convenience, and uniform distribution (obviously they don't use uniform distributions, as bodies are almost as white as engines most of the time, while pavement/road is way different sometimes, but just for ease let's say it is uniform), so at night 2 degrees per shade difference, day 1 degree. If you have a body temperature of 36 and background of 20 degrees, the difference at day is 16 shades of grey. At night, with your background at 4 degrees, the difference is also 16 shades.
With high sensitivity / better sensors and processing power / better hard+software, I don't see why today's high-end FLIR systems would produce a worse image at day than night to the point where it affects targeting. I think we're way past that and now it's more about resolution/range and maybe additional processing to produce more depth in the image.
That said, I'm out of my depth here...hehe