It is my experience that while there might be some inaccuracies in a meter, especially with respect to very low resistance measurements, we are generally speaking of small percentage tolerances. The greatest factor that gives me pause and reservations on my measurements is always seemingly fluctuating readings that are hard to pin down. But the reason is very simple.... all metal eventually develops an oxide coating that is invisible but affects readings. If you've found you get one reading when you initially place a probe that is suspect, but you press harder and you get a different more stable reading, this is almost surely the reason. This happens on even very expensive meters such as Fluke. While the probe tips has chrome tips that looks beautiful, the oxide coating is real, even if invisible. Basically, I just clean them every so often with plastic polish compound and then the oxide coating is wiped away and I get very stable readings. I also clean them before using them if they've been sitting around for any appreciable amount of time. Of course the component leads are also prone to this oxidation too, so the components under test will need to be cleaned for precise measurements. This could be done with some isopropyl and fiberglass tipped brush, or whatever you have since fiberglass tipped brushes are not common in most folks tool arsenal.