I believe this would operate as you propose - but it has some drawbacks.
The lens is more expensive than the window.
I believe the “lens 2” you show is a more expensive optic than the existing one.
If the optics are not perfectly matched, the beam is diverges.
Even if the optics are perfectly matched, the beam is still slightly diverging, because of quantum mechanical effects.
(most importantly) if the beam reflects in the current design, it diverges so that the power density drops quickly. In the proposed design, if the beam reflects, it’s extremely destructive.
Aye, Working as a mechanical designer for years always took a degree of physics to make things work, but now it takes a degree in physics to even understand what you are doing.
I think it is probably impractical because the second lens would see the focused beam at maximum power density, whereas all the other optical components handle the beam at its minimum power density.
