TLDR: I think the internal temp sensor is fairly accurate - from one data point.
I decided to check the interior temp sensor on my car. External reading was 21C, internal 25C.
I played around with the climate control, and it seemed to switch over from cooling to heating with the fans going quiet at 25C.
There was no sun in the drivers area, so I opened both doors to let a through draft. After a while, both internal and external temperature sensors are reading 21C.
My weather station was reading 18.9C at the time, and a thermistor in the car (in the breeze) was reading 18.2C
An IR reading from the plastic area around the internal temp sensor was showing 22C - but I'm guessing on what emissivity value to use.
As there will be quite a lot of thermal mass in a car which was still partially in the sun, and had been in more direct sunshine earlier, the slightly elevated temperatures from both sensors could be expected.
From this one data point, the internal temp reading seems to reading correctly.
What I will do is check again later tonight when the sun has gone down, and temperatures have dropped. I expect the internal temp to read much higher than ambient at that time. But I'll leave the doors open and see what happens....