Op Amp selection for a minimal component clipper / limiter is my interest
Recovery Time is the parametric most indicative of good performance; unfortunately its charachterisation is uncommon
Constraints firstly: 5V single-rail operation, VFB internal architecture, audio frequency range, dual in standard pinout.
Key requirement: release time after limiting/clipping at/below the rails, 1st stage will take an ~1Vpp LTC (linear time code, 2k, 4k8, ... Hz content) and diff Rx it with again of ~5, 2nd stage will slice it using a comparator topology. A 3 rd party conceptual design shows the use of a TLC2271ID as working fiction. In the past I have seen similar processing done with TL072's.
Attributes such as RRO and RRI may assist, but really it is the recovery from the rails of drive saturation that matters.
The output of interest is the timing of the zero crossings, their nominal minimum spacing is 1 / 4.8 kHz i.e. 208 us, so the release time requirement is modest < 1 us VG, < 10 us tollerable. However, less jitter should enable the design of a more discriminating detector.
I am discounting the use of zenners and biased diodes in the feedback path, the classic precission clipper designs, too many parts for the BoM and you can't get zenners with a sufficiently low zenner voltage.