I think here the case is simple. Broder did not record his speed accurately, he just drove "around 45 mph", which very well might be actually 55 mph, on a good road with a good car 10 mph difference in speed doesn't feel at all. However, he can't just say "I didn't really collect the data accurately" since it'd hurt his credibility, so he mentions completely unrelated discrepancy in hope that most of the readers would say "ok, it's plausible, maybe he's right, maybe he's not, we're not engineers, we don't know whose fault it is". Thus both Tesla's point would be neutralized and Broder's credibility won't be hurt.
That is a possibility too. On a mass-produced car, I'd evaluate this as pretty low probability, since setting speedometer for the actual car specs is a basic expectation and if any company would let wrongly calibrated cars out of their factory it would be a huge liability issue and a very loud scandal. However, if custom modifications were made to the car, it is not out of the question, theoretically. But then the effect of the smaller tires should be the opposite, not? I.e. with smaller tires the speedometer should overestimate the speed, not underestimate?
