While I wouldn’t say I agree with it, I’ve heard the argument that high specialization strongly selects for a student’s technical proficiency. As a PhD student myself, I’ve seen examples of this phenomena with highly intelligent individuals leaving our program as they couldn’t foster the drive to learn the requisite highly specialized technical skills. In contrast, I’ve also met individuals who developed their technical skills with rabid obsession, but who had trouble grasping the general significance of major publications in their field when the paper didn’t fall exactly within their own line of research. The second type of individual makes for a wonderful post doc or research scientist, but only until the techniques they’ve master are automated or funding for their specific lines of work goes dry.

Still, I’d say the best researchers/professors that I’ve met combined a deep understanding of the techniques in their own line of research with a broad understanding of their field and adjacent fields. Personally, I’d fear that removing technical proficiency as a requirement for a PhD would eliminate an essential bullshit detector that filters out certain people; people who do a great job speaking in broad generality but are incapable of formulating a path to accomplish these grand aims.

(As a quick anecdote, I once saw a research presentation from a grad student who made broad claims about his project’s direct role in curing many forms of cancer. As he progressed, we learned he was only exploring strategies to synthesize a potential ligand that some other researchers would consider using in imaging a specific cancer within rat models. To this end he showed very little progress, but continued to speak at great lengths about the imminent eradication of cancer as he concluded his talked.)