There are legitimate reasons for these kinds of designs. For example, oscilloscope manufacturers often have the exact same hardware for different models with wildly different pricing, just having different software enabled or not. Why shouldn't a multimeter manufacturer do similar (using slightly different hardware instead of enabling software features) for different models of meters at different price points? Another reason for having a design support different components might be for flexibility in parts supply (which might be seen as more important now after the COVID parts crisis).
Sure, Fluke does it with their small 113-117 series. Some of the models share a common PCB and they will have unpopulated footprints on some version. For example, there will be an empty place for "PTC2" in the models without Low-Z. If UNI-T had two models, one with CAT ratings and one without, then omitting the MOVs and putting in smaller fuses, etc might be a similar deal. However, when they put the SAME CAT RATINGS on the meters and then include or omit the MOVs, big fuses, etc depending on whether the meter is being shipped to a jurisdiction that actually checks vs one where they think they can get away with leaving them out, then that's a bit different IMO.
Does anyone remember the fake airbag covers that shady used car salvage builders would use?
https://www.oig.dot.gov/library-item/31564It's like that, IMO. Screw the people that trust your specs, only the ones that know what to look for and check for it get what they are paying for.