I have two devices testing the 5Ghz wifi frequency.
When I swap between the two devices placing them in the exact same spot at the same time of day, the Qualcomm device reports a noise level of 105 dbm while the Mediatek reports a noise level of 91 dbm. Can chipset choice really affect noise level?
More the circuitry towards antennas.
Oh really? I have a TP-Link Archer C7 with the same Qualcomm chipset laying around somewhere. I'll test that one and see what results come back too.
Beyond the above, don't expect the values displayed to be exact or calibrated - reality might differ a lot from your readings.
Update:
I tested my Archer C7 V5 and found it also reported 2.4 GHz noise of 95 dbm and 5 GHz noise of 105 dbm which means both Qualcomm devices get the same results. Based on what was suggested above, this would suggest that the Mediatek router might have a weaker radio circuitry design or components?
Not necessarily. As already mentioned, these values aren't calibrated - at all. You're are grossly overestimating their meaning, but the proof is in the pudding - and that you can only determine by eating it benchmarking them against each other.
I also get the same values on an Archer C60 v1.