Good ol' thermostat testing
I am not trying to advise you to use a certain thermostat. I am merely stating my data so you can make an informed decision. Crawford Performance says they use the OEM thermostat since it ensures the properly operating temperature and that all the engine internal clearances are within spec due to that proper temperature. Also, Koyo designs and tests their radiators using the OEM thermostat.
My testing was far from laboratory scientific. I took measurements off of the AccessPort in live data monitoring mode. Readings were always taken in the afternoon on a sunny day, upper 80s to mid 90s. I tested stop and go traffic, city driving, and highway driving. I drove off boost and went with the flow of traffic; this is not WOT testing, just casual acceleration and cruising.
stock thermostat (with stock radiator, coolant 30k old)
I wish I would have taken more measurements on the stock setup. City driving was mid to upper 190s. Some temperature spikes got as high as the mid 200s. Highway driving was in the low to mid 180s. The temperature spikes were rather interesting. I was stage 2, but I was driving off boost, very easy, and I would think that the stock system should have more cooling capacity during city driving. I guess that's what the fans are for. WolfPlayer posted a snippet of the manual, which said the stock thermostat begins to open at 172*F, but doesn't fully open until 196*F.
GMS 170*F thermostat (with Koyo radiator, shroud, hoses, fresh coolant ~55% water)
195*F was the go to temperature. City driving, highway driving, even when it was dark and cool at night. The temperature was rock steady. At red lights, it would get up to ~201*F. In stop and go traffic, it would get into the high 200s and the fans would turn on.
I determined that the 195*F was the result of the thermostat. I could pull up to a red light at 195*F and would wait until it was ~200*F. At this point, the thermostat was fully open due to the lack of air flow. When the light turned green, I would accelerate to 45mph. In the process, the coolant would drop to 185*F. As I drove down the road at 45mph, the coolant slowly rasied back up to 195*F and stuck there. I sent a PM to GMS on NASIOC of why a 170*F thermostat would always like to go to 195*F. No reply. Because a 195*F temperature seemed too high for a 170*F thermostat, I assumed it was defective, but read on...
GrimmSpeed 160*F thermostat (with Koyo radiator, shroud, hoses, fresh coolant ~55% water)
I was not happy with the results of the GMS thermostat. Too hot for my preference. I then went with a tried and true aftermarket thermostat - GrimmSpeed. I put new coolant in, making sure to keep the same ratio (1 gallon Subaru coolant, a bottle of coolant conditioner, and distilled water to top it off.) The results are staggering. The go to temperature is now 176*F for both city and highway driving. That's what I expected out of a 160*F thermostat. It makes a big difference pulling into the garage with coolant at 176*F instead of 195*F. At red lights, the temperature is very dependent on how long the light is. Usually, the light turns greens when the coolant is in the low to mid 180s. During stop and go traffic, it has just the cracked 190*F barrier. I have yet to hit the 195*F that the GMS thermostat always used to run. The OEM coolant gauge in the cluster now sits a hair over the 1/3 line. In stop and go traffic, I'll see the gauge in the "normal" position that it used to be at during the OEM and GMS thermostats.
It's a conspiracy!
The GrimmSpeed and GMS thermostats are both made by the same company - MotoRad.
"Inconceivable!" says the guy from The Princess Bride. MotoRad fail safe thermostats... So, if they're both made by the same company, the lingering question I have is why a 170*F thermostat kept the coolant at 195*F. By switching to a thermostat that had a 10*F colder opening temperature, I dropped the "go to" temperature by 19*F. That just doesn't make sense to me.
GrimmSpeed 160*F on the left, GMS 170*F on the right