I'm not sure how you consider significantly less than 16C ambient a hot day?
What you want to look at is the difference between ambient and measured temperature. For all intents and purposes in the realm of what we're doing here, temperature differentials will stay the same across a range of temperatures. If your ambient is 10C, and you're running at 16C, then when ambient is at 40C you'll be running at 46C. Granted this isn't exact, but it's not a bad approximation.
This is assuming the airflow over the computer is equal to what it will be in the car. You should also do these tests running something like Prime95, or some other equivalent stress test application which will ensure maximum heat generation.