I am not from the NL but I have moved here to attend a MSc program (Computer Science department), I am not pursuing a PhD myself but I know a lot of people that do. I think that the system has its pros and cons, as always, although I don't agree with the part about not choosing the topic because (as also other people pointed out) policies can often vary from supervisor to supervisor, and in any case a student basically chooses, more or less, the boundaries of his/her topic when he/she decides to which position apply, positions that always have fairly clear project descriptions and include contact details of professors that are generally open about clarifications. Also, even if it is formally difficult to change supervisor, in fact if a student decides to work more with another professor for any reason, the original supervisor does not usually raise any practical problem.

However, a big problem that I have experienced and seen is that PhDs negatively impact the quality of teaching. Master programs in the NL aim at being very high level (often succeeding), offering also courses on specific advanced subjects. Usually, PhDs have no control over what they will teach, they are simply allocated as "resources" by the faculty, even just to cover a minor part of a course or to supervise/correct practical assignments. This leads to ridiculous situations with courses that, for example, teach advanced rendering techniques but for supervising the practical part allocate a PhD candidate that has clearly never ever rendered a triangle on a screen in his/her life, and who had no time to properly prepare beforehand. I think this is a visible "weak point" of the whole system, and it should be addressed, somehow, also because I have noticed it being treated as some kind of "taboo argument" in the past.