To ST3 it does not really matter which format you use. The audio engine uses internally 44.1 or 48kHz at 32bit floating. So anything up to that point might make sense. Anything above that (for example 96kHz) is useless as it will be downsampled to the internal sample rate.
Peak and distortion are not related to your sample rate. If you clip your audio signal you will get distortion with any sample rate and if you don't, any sample rate of at least 44.1kHz will sound basically the same. Especially in a venue with suboptimal acoustics listened through a PA system.
The biggest difference is that higher sample rates will allow higher frequencies (44.1kHz will not allow any frequency above 22kHz, with 96Khz you could theoretically get up to 48kHz frequency). In practice this is upper limits are a bit lower due to dithering.
And regarding the bit rate: It makes sense to record signals with at least 24 bit to have a high resolution even for low volume signals. A 24 bit signal has a signal to noise ratio of 144dB, whereas 16 bit audio has a SNR of 96dB. The difference might sound huge, but in applications like ST3 where the audio files are at an optimal level, the 16bit resolution is already plenty sufficient.