I think this approach briliant. However, I believe it can be simplified to avoid depending to trigonometry and geometry series, while keeping the original idea. I have no idea if it is similar to any existing proof of the theorom, but I put my thought into a post [1].
There really isn't any trigonometry here apart from sine rule, and because sine rule is derived from similar triangles, anything you can say using sine rule, you can also say using similar triangles.
The trigonometry thing is simply a marketing gimmick for this proof. There is no more or less trigonometry in this proof than there is in Einstein's proof. In fact, you can just taken Einstein's construction and reformulated that proof in their language by using sine rule instead of similar triangles. But then the gimmick would be too obvious.
I like this as a proof better than the OP. Not only does it avoid the sin2(a)/cos2(a)=1 pitfall but shows clearly the math without use of trig. This part is the yummy bit: >Since a<b and a+b=90 degrees then a<45 degrees, and therefore 'BAD' < 90 degrees, which means that if we draw a line 'l' that is perpendicular to 'BA' at 'B', 'l' would never be parallel with the line 'AD' and they cut each other at a point 'E' like the Figure B.
[1] https://eneim.notion.site/On-a-new-proof-of-the-Pythagorean-....