You won't get any benefits from rendering your files to 32 bit float prior to processing. Every digital audio processing step, from complex plug-ins to tasks as simple as summing, have their own -higher- internal processing bit depth and (over)sampling frequency, and convert your files on the fly to and from them (for instance Samplitude internal mixer work with 32 bit float while Pro Tools work with 48 bit fix, Fairlight Crystal Core uses 36 bit float for the mix and 72 bit fix for the Eq, etc). This is true for almost every piece of software and hardware meant to process audio.
Using files with a 32 bit float bit depth will mostly result in (way) bigger files, more processing time and an additional unwanted dithering step when converting back to 24 bit before delivery. All of this whitout benefits.
So if I were you, I won't bother.