IVR pulse at 4 volts gauges read low

As said, part of the problem is that a voltmeter will not properly follow the voltage pulses. Even if the IVR is swinging to +12v, it is for too short a time for the voltmeter to typically show it. Occasionally, a digital voltmeter might show the true pulsed IVR voltage if it just happens to sample it at the right instant, and has the right meter response. The best you can do is to watch it for a minute and note the highest readings it measures, and consider that the pulsed IVR voltage. You need an oscilloscope or similar fast voltage-reading device, or an old, slow thermal voltmeter, to really know.

BTW, you CAN measure this properly by just disconnect the temp sensor wire from the temp sensor and measuring the voltage from the temp sensor wire. That connects to the IVR output through the temp gauge's resistance but with no current through the sensor, the gauge resistance's voltage drop is zero (i.e., it 'dissappears'), so the wire shows the actual IVR voltage. You don't need to disconnect all 3.

The IVR voltage being 'low' is indeed consistent with the low gauge readings. It is more likely that the IVR is 'short cycling', i.e., the 12 pulses that it is putting out are too short and giving you a low average voltage. If your Fluke meter is not showing anything, then I really suspect the short cycling of the IVR. The Tripplite analog meter is showing all that it can....

Getting to the IVR in your case is a pain; otherwise there is an easy way to check with a 6V battery.