I don't. Another audio quality myth is that 24-bit audio will unlock some sort of audiophile nirvana because it’s that much more data-dense, but in terms of perceptual audio, any improvement will be lost on human ears. Capturing more data per sample does have benefits for dynamic range, but the benefits are pretty much exclusively in the domain of recording not human listening.
I agree: 16-bit resolution gives 65,536 values to represent samples. 24-bit resolution gives 16,777,216 possible values. This is huge improvement and since most (probably all) plugins are using floating point arithmetic, which has limited precision - it makes computation more precise if you start doing it from more precise values. You can export final product to 16-bit and you probably would not hear much difference, but recording and processing in 24-bit depth is a must, in my opinion.