A shoulder is a very different obstacle than the patellofemoral joint. If the ortho suspected meniscus damage, hemarthrosis, tumor, whatever, the MRI would be the way to go. However, the patellofemoral joint doesn't have any of this. It has articular cartilage and possible plicae (that aren't even well viewed in MRI, I don't believe). It's a completely different structure. For the shoulder, there's a TON of stuff that can't be visualized on x-ray that can be on MRI - labrum, muscles, tendons, capsule, etc. The PFJ is a little bit more limited in structures.
The issue with PFJ is that it isn't solely an exercise treatment. It should be (ideally) a lot of hands on. Patellar maltracking can be caused by having weak hip external rotators and abductors (which someone can exercise), but it can also be due to tightness of the lateral structures of the knee. That tightness requires soft tissue work, joint mobes, etc. We also assess for patellar deformities and femoral deformities that may be contributing so we can estimate the potential effectiveness of PT. PTs don't (or shouldn't) just do exercise. And there's a lot of research on given exercises directed toward activating certain patterns, muscles, etc that people don't realize. Status post labrum repair, I understand why you have that viewpoint because most of PT is just getting back into moving and strengthening yourself. With a chronic issue like patellofemoral dysfunction, there's a lot more that has to change.