The best mastering chain ends up being nothing but a set of ears and a thumb's up.
But, since that almost never happens...
I may switch positions between different eq options and my own choice of tape and saturation- either in or out of the computer.
I go for a combination of preamplification tubes and tape to choose some soft increase of the signal without change to the transients, which does some of what most people might try to gain from the limiter.
The limiter is usually last in the chain other than any dither options, because it is designed to prevent overs, but also because it is specifically designed to handle the timing elements of the music from slow, medium, fast, simple, to complex, with a certain amount of finesse and transparency. The compressor, especially depending on music style and genre, can help lock the rhythm of the song into place, so eq>compressor>limiter, with tube or tape either before or after eq.
Also, I sometimes prefer to use two instances of limiter or 2 or 3 different limiters at the end- letting each one do less work on the peaks- this gets a better balance of exactly how I want each one affecting the full chain.
Like others say- the best end result is already happening in the mix.
I personally do have a system for my own work and sometimes for stems if they are done similarly-
I have a simulated mastering chain when mixing that I turn on an bypass, and decide which instruments to change the volume or eq in the mix based on what it is likely to need once all blended together. This is a difficult relationship to define in words, but I hope to do a tutorial series or audio tut some day to define it better. The general idea is that there is more than one stage that can happen during mixing. going from completely dry signals coming in to mixed to mastered may benefit from a 2-stage mixing process- taking dry tracks to polished and rendered individually treated tracks, then blending them in a mix while A/B'ing a simulated mastering to make smaller eq/gain choices.
Simplest answer to that part- sometimes we over eq or overdo the difference in gain from one part or another because of what is already there, but if we make some better adjustments and deal with harmonics, problems on each track before the full mix, it can actually boost the quality of the mix, and thus give even better mastering options!
