A PhD is intended to demonstrate that a person is capable of pursuing a course of research within a field of study, and has a mastery of the existing knowledge and pertinent skills/methodologies.
Unfortunately, in some fields, the PhD has sadly become a form of lazy gatekeeping, and reimagining the PhD in those cases may just be a matter of waiting for the old guard to step down.
In some other fields, the PhD is probably ripe for reimagining, but "independence" might not be the right lens through which to think about it. For example, it would be exceedingly difficult for a would-be physicist to independently build a large hadron collider to do their research with. The institutions that make those sorts of resources even possible are inherently large/complex, and in those cases the PhD is a less sad and lazy form of gatekeeping (i.e., you need to do a set of progressively more complex stuff under the tutelage of someone more experienced before we let you touch all of our big expensive things).
In yet some other fields, which may be younger, have less of an established orthodoxy, and lower overhead to participate in (e.g., the economic/anthropological open source software type work the author does), it seems totally reasonable that a person could develop and demonstrate their knowledge, skills, and abilities on their own and without a formal course of study. In such fields, the concept of the PhD doesn't seem as ready for reimagining as it does retirement.