dada2 alone is more appropriate, but there is nothing technically wrong with subsequent clustering.
Yes, those same settings are used in the denoise-pyro method. This was written by the dada2 developers themselves — you should check their website for any benchmarking (I have not tested this method personally since I do not use pyro data).
In theory dada2 should provide the same type of output for pyro data, and subsequent clustering should be unnecessary.
I hope that helps!