I think this is a bad argument. It's like saying only because someone clever choose path A it is wrong to choose path B.
They still might learn from their own mistakes, but at least they are trying something different. I'm not sure vaporizing tin-droplets can be improved enough to solve all of ASML's problems.