[jakeru]

Since the tie rods in our macpherson strut equipped cars are located behind/aft of the axis between the ball joints and the strut mounts, and vertically a little bit higher than the ball joint (since both the ball joints and tie rod end are designed to fit inside the round wheel barrels, the tie rod end is higher to work from packaging perspective), trigonometry tells us that moving the upper strut mount inward would also require the tie rod end to move a fraction of that distance inwards in order to maintain the same toe setting. If tie rod were left at same position, moving the upper strut mount inward would cause toe-in.

It would be the opposite effect, however, if the way you accomplished changing the camber was to move the lower ball joints outward (e.g., using adjustable a-arms).

If would also be the opposite effect if your suspension design had the steering rack in front of the front axles. Not sure how many cars have setups like that, but I'm sure there are a few.

I remember a fellow autocrosser would adjust his toe before and after each event, on site... he had a homemade, I think substantially wooden, trammel bar he set up across the tires/rims, and adjusted the tie rod end at only one side of the car (for speeding up making the adjustment vs adjusting at both sides). I don't think he was adjusting camber, though! I think he was dialing in some additional toe-out for the event, and then reverting to conservative, tire-wear friendly toe settings for the street trip back home.