32 bit floating adds NO more bits to 24 bit integers. In fact, below 0db, there is NO practical difference.
The ONLY advantage of 32 bit floating is that it is impervious to overloads, which helps DAWs cope with a wider range of users as gain staging is not as critical.