What about including some actual measurements and tests from meters? We should have good data about some of the meters from this forum alone. I know it might be a lot of work, but cheaper meters like Uni-t always have their quirks. Like ut61e has high burden voltage because of the relatively high value resistors in current range, where as ut61d as example has low burden voltage. From what I've understood from EEVblog and TheAmpHour Fluke meters always exceed their specs, because all the drift of all the components has been included in the error calculations of the specs.
Specs quoted from manufacturers don't always tell the whole truth, for example ut61e can also measure hertz far higher than what they actually tell in spec.