Like the guys say, Differential pressure measurement is designed for DP. Subtracting one gage pressure reading from another can work OK if the readings and the difference are a good percentage of span.
There are mechanical DP gauges with (optional) switches (Midwest, Orange Research) and DP transmitters.
Years ago I actually saw two brand new, out-of-the-box 2" dial 0-60 PSI, 3% accuracy bourdon tube pressure gauges, one upstream, one downstream on a brand new filter. The gauge accuracy was technically ±1.8 psi (plus an undetermined amount of parallax)
With the new filter presumably dropping very little pressure the upstream gauge read on the low side at 42psi, the downstream gauge read on the high side 43 psi.
The differential calculation is "High side" minus "Low side" = 43 psi minus 42 psi = -1 psi, or reverse flow. It took awhile before there was enough pressure drop to indicate forward flow pressure drop.
Pressure transmitters are rated "percent full scale" accuracy, so if the readings/values are near the top of the scale, the accuracy is better because it is less as a percentage of the reading, than near the bottom of the scale where the error can approach the reading value itself, for instance, a pressure reading of 5 psi with a 1% accuracy transmitter ranged 0-500 gives a potential error of ±5 psi at 5 psi. An extreme example, but those are what the numbers shake out to.
You should calculate the transmitters' error at the expected pressure values and see how the plus and minus accuracy affects the readings you can expect.