Doubts on Tesla/Fisker

Sure, for some things there is no substitute for physical testing (fabric durability would be a good example). But it’s much less compared to what you used to have to do, where you’d have to iterate a design multiple times based on test results. Now in most cases you can just do validation testing to make sure your simulation results are good. Tesla does a lot of testing of complete cars in the preproduction phase, since those component interactions are difficult to simulate. But torsion testing an axle on a test rig? I’d be surprised if they bothered. I’m sure Ford has a whole axle test building somewhere filled with dynos and a staff of operators.