"qaac" reportedly uses 32 bit float internally where as "fdk" uses only 16 bit integer.
Now lets put aside the fact that one of them might be more efficient than the other. So when using a 24 bit source is there a reason to use "qaac" instead of "fdk"?
Is doing 24 bit -> 16 bit as much of a sin as transcoding?
Answer is it doesn't matter:
24/192 Music Downloads...and why they make no sense (https://people.xiph.org/~xiphmont/demo/neil-young.html)
24 bit vs 16 bit: Nobody can hear any difference in the "music". 24 bit is for audio engineering and the like. There was never produced a recording that demanded 24 bit for faithful reproduction.
If you are already using qaac, fine. The 24 bit audio is internally changed to 32 bit float so no loss there. Just stay with it.
If you are already using fdk, 24 bit audio will be converted to 16 bit and you will never notice it. Just stay with it.