That could work, sort of. The beauty of the Eckel circuit is the substantial range of its overlapping simultaneous outputs, as you go through range changes.
You won't be trying to digitize an opamp all the way to its saturation, and wait for the next stage to begin working.
If you have a very noisy signal, having two amplifier signal versions, at different gains, lets you process data from two channals and get a clean result.
Note, it has a similar parts count: one opamp per range, and JFET instead of a diode. It'd be a good addition to your bag of tricks.
Also, Eckel's circuit can work through zero, with bipolar input currents.