I noticed the temperature thing here last night, and I'm in Florida. Not exactly brutal cold. It was low 60's. Maybe 16 degrees change.
I think we have to account for ambient temps, what temp the initial build was tested and set as a new coil at. THEN the biggest thing is the air coming IN and changing things. Blasting 30F onto the coil cooling it is a lot different than 65-70F when in the house. The differential. The average ambient temp in my house is 77-78 with the air on

So now the wattage and temp setting you tested and found a happy medium for setting to stop protection from constantly kicking in, seems to go out the window once your incoming Air changes and the ambient temp of the device.
I also noticed the opposite sitting on the hot porch today that was probably 90. Even the 10-12 degrees was noticable in hitting protection sooner. (Kayfun Lite 2) with bleeds turned in a bit.
I had 3 builds going. 2 kayfun Lite v2's and a Russian clone. All were 1.4-1.6, 30 ga at about 3mm, 12-13 wraps both spaced and pressed.
Today I sat and thought about it. I wanted to try a higher ohm and also try and get the mass of the coil up a bit.
I'm at 4.3mm ( 11/64 drill bit ) and a contact coil that was heated and pressed. Came out as 0.24 on tester and 0.23 via the dispay. 12-13 wraps of 28ga. At 12 watts I could just hold the fire button. The coil was glowing nice and even and at that temp it was at 1.24 Ohms. Nice glow. Legs looked great.

I have it set at 20w and 430F. I can vape on it 3 or 4 times, in the house, mouth-lung in a row and not hit the flashing temp protection. Warmer vape than same temp on 3mm 1.6 setup. Walk outside with it about 10 degrees cooler and do the same.. Flashes.
They had to of thought of this. There is a temp sensor on the board for overheat. If they used this for calculations it'll never be cood enough unless you walk around with your mod in your hand and the case gets cold enough.
I may be totally wrong and I'm throwing this all out there for discussion. But I keep thinking temperature differential. As in ambient air being heated or cooled. In this case a coil being cooled by air that changes in temperature. Just like your car AC might despense ice cubes at night, but during the day you're wishing it could cool enough because of the ambient temp at the condensor in front of the rad. You can only get so much of a differential. Maybe the cold air entering and hitting the coil is effecting the detection of the temperature. Keeping the coil temps down and the wattage goes higher trying to reach it's "governor" point using the "new coil" temperatures that were saved.
I'm going to keep testing larger diameters and hopefully give it more mass/ohms to work with.