After I got autoguiding, I ran similar experiments numerous times. I found that for my location, the optimum exposure length varied from night to night. On one night, a single long exposure would be strikingly better, but more often than not, a single long exposure would be inferior. It depended on how steady the conditions were. Long exposures can be wrecked by a long list of disturbances; unsteady atmospheric conditions, occasional wind gusts at ground level, clouds/high level haze that come in and out, jet planes, satellites, someone driving by, pets,,,,you get the idea. My general rule is to use the longest exposures that conditions will permit. This can range from 2 seconds to 5 minutes.
It also depends on the camera and the focal length.
My starting point is usually 10 seconds for the 183 or 290, and 30 seconds for the DSLRs.
The DSLRs benefit from longer exposures. So, in their case, I tend to run longer exposures and use Astrotoaster to select the best exposures for stacking. AT gives each exposure a score. You tell it what best percentage you want to keep. On a good night I might set it to keep 90% of the exposures. On a not so good night, I might set it to 40%. AT also lists individual exposures so that I can click through them while it is stacking and delete bad exposures on the fly. Then I can restack at the end. This is a good way of getting rid of satellite traces, etc.
Sharpcap can also be used to screen out bad exposures, using the filters, but not quite as well as Astrotoaster, in my experience. When using Sharpcap, I usually get the best results by shortening exposures.
I have dark skies, so I don't have to worry about long exposures saturating pixels with sky glow. But, as pointed out above, some people do.
Anyway. That is what works for me. Your situation may be different.