I've seen plenty of wiped cam lobes, recessed seats, and a number or cracked exhaust valves in motors I've torn down. I have seen zero instances of any of those in motors I've owned, built, or tuned for customers.

There's no history I know of cracked valves in high performance motors running cams with much longer duration than stock, or stock cams running greatly reduced lash. There's plenty of seat time for heat transfer. IMHO, the primary cause of overheated exhaust valves is pinging.

You can't tune a motor unless all the cylinders act identically. Valve lash has a big impact on duration (not lift, I agree). If your adjustment method does not result in identical duration, the best you can do is tune for an average of what the individual cylinders want, and none of them will be optimized.

I can't say anything scientific about valve train flex under static testing versus actual running. I can say, with a high level of assurance, that motors I've tuned using my quirky methods run a lot better than when they arrived. The change is worth the added effort -- I can't think of a customer who hasn't commented on the improvement.

Sorry if I come off like some sorta smug know-it-all, but I really am trying to pass on some information I think is worthwhile.