One of the things that fascinates me about academia is the diversity of training environments and philosophies. Whereas some labs are genuinely run by people who should not even own a hamster (let alone be responsible for providing a good environment for the development of a scientist and person), most PIs have a method that they feel is effective for getting their people to where the PI believes* they need to be.
In the last 6 months I have been an examining member on three different thesis committees, all with very different levels of PI involvement in the final product. Knowing the different PIs, I can say that each is engaged in the research their students are doing and meet regularly with them, but each to a very different approach to the student's thesis and presentation.
The lifeguard: In one instance it was clear that the PI had been through several thesis drafts with the student and made significant editorial changes along the way. As a result, the thesis was a very complete document without much for the examiners to deal with on that front. Interestingly, it would appear that being too well put together can have its downsides, as one committee member wondered aloud to me prior to the exam whether they should be congratulating the student or PI on the quality of the thesis.
The mildly attentive parent: In this case, it was clear the PI had a hand in the final outcome, but that the writing was largely the student's work. The committee had a bit more to focus on, in terms of the editorial side, but I felt like this was better received than the thesis mentioned above because the perception was that there was more ownership from the grad student.
Good luck with that: On the other side of the spectrum is the PI who let's the student write the whole thing and send it out without the PI providing much or any feedback. As a committee member, these are the ones you fear. Sure, maybe there is the exceptional student who writes science like they were born doing it, but this is not the norm.
Obviously if the student has published several chapters already PI input becomes less of a concern because it has been invested along the way, but it is common in my field for students to have some chapters that are not submitted for publication just yet. Whereas I can see the perspective of wanting the students to stand on their own two feet at the end of the road, I think my own philosophy tends more towards the middle. Just because someone is earning a graduate degree doesn't mean that their training as a communicator of their science is done.
Or perhaps PI involvement is just a function of career stage.
*A view not always shared by the trainee