Ok I decided to do some experimenting on batteries to see what the ACTUAL effects are on soldering wires to a 2032 battery, I've soldered to various button cells over the years during many of my electronics tinkering projects so I feel qualified to do this.
Equipment used:
Weller WESD51 Soldering station
Amtech flux
Radio Shack .062 diameter ROsin core solder (I'm out of my usual Kester)
Rubbing Alcohol
Basic Wire
Extech DVOM
2x Sony CR2032 Lithium cells
100Ohm Resister (to load the cell while testing)
Voltage checking method
Attach 1 end of the 100Ohm Resistor to an alligator clip and hook that to the test probes of the meter then test battery, this will provide a (roughly) 30 milliamp load to the battery while testing
First Battery - Double soldering technique (tin and then tack)
Starting Voltage 3.09v, Measured once a day for 3 days, 3.08-9 for all 3 days
Method :
Sand battery surfaces lightly to scuff
Apply Flux
Apply light amount of solder to battery keeping iron contact to UNDER 1 second
Immediately apply a paper towel soaked in alcohol to cool the battery
Wait till FULLY cool
Repeat on other Terminal
After this I tested the voltage again, 3.07v
Tin wires to be used
When totally cool, tack wire to positive terminal (Again total iron dwell time under 1 second)
Alcohol to cool
Wait for total cool
Repeat for Neg terminal
After doing this and allowing the battery to cool voltage tested at 3.27 again
10 minutes later, voltage read 3.09
10 Minutes later, 3.02
2 Days later 2.4

3 Days later holding at 2.4
As we can see the voltage was pretty stable PRIOR to soldering the leads, once I attached the leads it was lower and then dropped off sharply over the next 2 days....
Second battery (tin and attach at the same time)
Starting Voltage 3.07v, Measured once a day for 3 days, 3.06-7 for all 3 days
Method:
Sand battery surfaces lightly to scuff
Tin Wires
Apply Flux
Tack wire to battery keeping iron contact to UNDER 1 second
Immediately apply a paper towel soaked in alcohol to cool the battery
Wait till FULLY cool
Repeat on other Terminal
Voltage after cool 3.05
10 minutes later, voltage read 3.05 (Hmm, Promising )
10 Minutes later, 3.05 (Looking good!!!)
1 day later 3.01 (Lower, but still good)
2 days later, 2.7

Still holding at 2.8
So as we can see it held better, but then within 2 days had a sharp dropoff in voltage, however it seems to have stabilized after that... I can only assume that method 1 because it heated the battery twice caused more damage hence the lower voltage out of it
Now I am loading these things with (Approx) 30 millamps which I believe is more then the TPMS would use as I cannot find the data sheet on them, I also don't know the dropout voltage, but I imagine they would work on these batteries for a time...
However, given the voltages observed, I am forced to conclude (by my totally accurate and perfect science methods

) that even when we do the change this way and the TPMS works, it is on borrowed time
Once the batteries that have the tabs on them arrive I'll repeat this test but soldering to the tabs this time and see if it's better...
