The exception is ESR. Components in parallel will only reduce the ESR, so an ESR reading of 10 Ohm across a fairly large cap is generally a good indication that that cap is bad. In troubleshooting you generally only care about order of magnitudes change in ESR. LCR meters (not just this model) are generally not the best tool for in-circuit ESR measurement. Dedicated ESR meters like the Bob Parker design tend to do a better job for in-circuit measurements.
Yes, you are right but that is not the problem. If there is a low impedance in parallel then you measure a low ESR. So a low reading is not a guaranty that the cap is indeed good. Most dedicated ESR meters do not measure ESR but impedance.
I have done alot of comparing in circuit and out circuit measurements and I think in situ measuring just is a waist of time. I only use it for some quick checks if a cap is hard to desolder.
For good in situ use he output must be < 0,3V because switchers are stuffed with shottky diodes. Bridges can alsop measure C and DF in situ because amplitude is adjustable. DF is a better parameter as ESR because ESR an sigh is meaningless.
I have a IET DE-5000 and I compared it to standard capacitors and inductors and measurements made with a Boonton 63H (200 pH resolution) and a GR-1620 ( aF resolution) , scope measurements and VNA measurements (RF-IV mode). I never compared it to the mastech and I do not think it is better because it is more expensive. But you pay for the name IET and therefor you know it will be good. They do not want to mess up their reputation.
The chipset alone is just a part of the meter. That does not guarantee it is good. The external components (jacks, resistors, pcb quality, testleads) can be garbage, de pcb design bad, the psu not good (or kill your meter) , the soldering done by amateurs etc, You can be sure this will be good if you buy the IET.
And what you buy is up to you. A Ferrari engine in a Kia does not make it the same car as a testa rosa