When opening those files in batch via Macro and applying effects such as pitch/bass/etc and finally exporting (all done automatically via Marcos pre-programed wonderful feature)
This is your music and your project so I’m not trying to tell you what to do, but usually mastering (or re-mastering) requires quite a bit of careful listening and human interaction, so a few additional seconds at the end shouldn’t matter. Usually, you’re going to listen-through the whole thing before starting (diagnosis before treatment) and then listen again when done. So, you’d typically spend at least 10 minutes on a 3-minutes song, or maybe longer depending on what needs to be done.
A good mixing/mastering engineer will also do a final-check on headphones an in their car and/or on some “downgraded” monitors, etc. That eats-up more time. A good amateur mixing/mastering engineer who doesn’t have great monitors or a studio with good acoustics will check the mix on everything they can get their hands on. This stuff is time consuming!
the file is downsampled from 24bit to 16bit and does not retain it’s default and original bit-depth or sample-rate.
FYI - By default, Audacity converts everything to 32-bit floating-point when you open the file. (There are technical advantages for using floating-point for digital signal processing.) The conversion from 24-bit to 32-bit floating-point (and back) is mathematically lossless.
It doesn’t get downsampled unless/until you export to a lower bit depth.
I kind-of understand you wanting to keep the original bit-depth, but “CD quality” is better than human hearing so there is no practical advantage to 24-bit audio. The guys who’ve done scientific, blind, level-matched, [u]ABX tests[/u] have pretty-much demonstrated that nobody can hear the difference between a high-resolution original and a copy downsampled to 16-bit/44.11kHz. Lots of “audiophiles” (and even audio professionals) claim that high resolution audio sounds better but usually they haven’t bothered with blind listening tests, or they’ll claim that blind listening tests are not valid, etc…
Even a good-quality MP3 can sound identical to the uncompressed original, or it can be very-difficult to hear the difference (in a proper blind listening test).
Plus, your 24-bit originals aren’t that great anyway or you wouldn’t be remastering.