If it is the gauge, it is easy to swap it out. But I don't think you have determined what is really going on yet. It could also be the sender or something else.
You haven't clarified if you are using a series 2 temperature gauge.
What sender did you purchase? I'd expect you to see some inaccuracy on the gauge even if the sender was working properly, as I'm not aware of any replacement senders matching the OEM series 1&2 sender. But if the sender reliably sends the same resistance at a given temperature, then you can figure out that when the gauge is reading "X", the temperature is really "Y". Then you just keep an eye out for whenever the gauge is reading lower or higher than the expected "X".
If possible, start by testing the sender and recording the resistance you are seeing at various temperatures. That can be compared against the data Tom H gathered when he tested the series 2 temperature gauge. You'll want to use an infrared thermometer to know the temperature, or remove the sender and use a pot of water to test it.
The temperature gauge can be tested, but you would need to use a variable resistor (potentiometer) rated to handle the necessary current.
A test rig I made to test multiple senders for DanR a while ago for his 2.8 V6 efforts:
Mike