[alocky (Andrew lockwood)]

If I might attempt to summarise:
Step1. Take as long an exposure as you can, stars nice and tight, no planes or satellites, sensor not saturating, particularly on the interesting bits.
Step2. Repeat step 1 until you are bored, or the target is lost in the evening sky.
Step3. Stack images.
Have a look at the sort of hours Martin Pugh spends on image acquisition- not to diminish his considerable processing skills, but that is 90% of the reason his images are so exceptional. Good data with high signal:noise.
It would seem for the vast majority of objects, 10-15 min subs seems to be a good starting point, with the exception of globular clusters. Looks like you need a range of sub lengths to capture the full dynamic range. Of course if you are doing narrowband, your subs will need to be longer. Not a game to be played with a poor mount.
Cheers,
Andrew.
PS I am not suggesting for a second that if you take as much data as Martin Pugh that your images will be as good. That extra 10% (my opinion only, others have suggested 60, after all only Mr Pugh knows for sure) does not come easily, you need to develop the skills to push the data as far as it can be pushed yourself. Fortunately, that part is mostly effort and not dollars.