I recently ran into this myself.
Make sure the gauge is working correctly first, just because the needle jumps up when you ground it is only a very partial test.
If you have yourself a potentiometer (2K would be nice, I used a 5K as it is what I had), set it to 1350 (note which way to turn it to reduce the value). Disconnect the plug from the sensor and connect the potentiometer between the gauge wire and ground (connect the center and 1 outer leg together on the pot for the 1 connection and the other outer leg for the other connection). Gauge should read 100°F - as you adjust the potentiometer to the lower setting the needle should rise almost right away (note it is not linear though).
On mine the needle didn't start to rise until I was around 200 ohms and would max at about 38 ohms - obviously wrong. Grounding the gauge would get the needle to max out like your test showed. On mine the needle obviously wouldn't rise as the engine never got hot enough to cause the needle to move (sender output never got below 200 ohms). Installed a different gauge and all was good on my car. When I tested the meter (I have equipment to do this) it showed that the 'reference' side of the meters coil was a lot lower than it should be (about 1/2) - it should measure the same on both sides, if anyone cares.
A quick test of this would be to know what mid scale resistance is suppose to be - install a resistor/pot and see if it reads mid scale. edit: now that I think about it, I seem to recall reading in the FSM that it gave the mid scale resistance to use (sorry I don't have access to my FSM ATM).
That being said, you can also test the temp sender - info can be found here:
http://www.gafiero.org/temp_gauge_repair.shtml