I don’t think there should be any variation, as most universities follow a very rigid structure and make sure with procedures that ensure that it is followed. No variation is allowed and the adviser and the committee makes sure it is followed. .For example I am giving the link to the guidelines of our university. The only variation may be from year to year there may be some small changes depending on APA guidelines (if there is any revision).

http://www.uni.edu/coe/sites/default/files/downloads/THESIS_MANUAL_Revised_July_2012.pdf

http://www.grad.uni.edu/thesis-dissertation

In my university, an article-based thesis is expected to consist of interconnected chapters (although, in some cases, these interconnections have not been very strong) and the author has to demontsrate this interconnectedness in a long introductory chapter and also in a relatively long results chapter. Using different methods is a plus, but using the same method is also allowed. It is also allowed to defend a monograph. It is not always possible to foresee if a PhD student will become a scientist in the future or select another career (in teaching or outside the university): all are expected to follow the same rules.

I don’t think there should be any variation, as most universities follow a very rigid structure and make sure with procedures that ensure that it is followed. No variation is allowed and the adviser and the committee makes sure it is followed. .For example I am giving the link to the guidelines of our university. The only variation may be from year to year there may be some small changes depending on APA guidelines (if there is any revision).

http://www.uni.edu/coe/sites/default/files/downloads/THESIS_MANUAL_Revised_July_2012.pdf

http://www.grad.uni.edu/thesis-dissertation

Dear @Marcel, I think that  PhD thesis manuscripts of future professors and future researchers differ in organisation and content because science becomes more specific and specialized.

The path to be a Researcher or a Professor pass through similar destinations, however via somewhat different set of routes/inclinations. At times it is hard to separate them; at times the difference is quite distinct. This is also dependent on individual interest, assigned or agreed objectives, affiliations and mentorship (supervision, in case of Ph.D. thesis). The discipline and the conventional understanding around that domain also has a good role in organization and content -fundamental science, applied science, action research, field research, development research …

Supervisors might be 'specialist' or 'generalist' in a given field. Perhaps this will determine whether PhD manuscripts will be more or less focused?

Dear Marcel, I am happy to learn that exists such beautiful things like your above mentioned ones. Unfortunately here in Greece, for the moment, the only 'object' that really matters is the will and egotism of the supervisor, thus leading to many kind of suffers for the PhD student... I am here to learn...

The question ,of  dear Marcel,  requires from us "the participators"  to think deeply before answering it. Many old-fashioned scholars, like me, will expectedly show a "rigid" attitude refusing any changes in the style of the PhD thesis as we know it at present & as it was after WW2.

In the world of today, the readers of any lengthy manuscript or any lengthy article do not have the patience to cover all the details thoroughly in most cases. There is a widespread trend towards acceptance of more concise direct-to-the point thesis. I envisage this trend to become global. Assume that a certain researcher has 10 new contributions to human knowledge in his/her field. I do not think that it is a "crime" to present the "ten" in one or 2 pages directly after the "Abstract". The "introduction" chapter may undergo shortening. The PhD thesis may become 2 parts: the main thesis which will be examined & the details "presented separately" but not examined unless a need arises.

Our present mentalities cannot shape the future fully & precisely. The future generations may come up with better educational processes including the high degrees but it is hoped that academic honesty will triumph ultimately.

But then again, there is the question of supervision. Do the referees of articles presented in PhD manuscripts act as supervisors, or not? For instance, why not adding all referee reports as supplementary information at the end of thesis manuscripts to provide a reliable picture concerning the evolution of the thesis contents?

This is an interesting question. If the thesis is written by a pure scientist-researcher, it might contain more abstract formulations/reasoning. And, if it is done by a researcher in technology school like engineering or management, then it might be less abstract and more focus might be on practical implications. 

I think that the future will influence both professors and PhD theses. The reasons for that are:

• The introduction of computer science, and especially computer algebra systems, like Mathematica, Maple, etc. These systems convert theoretical e.g mathematics into a experimental-like science.
• The focus on biology and human itself will transfer mathematics and all science. 

These changes will influence the form and the methods of future theses. For example experimental mathematics, will give a proof that all bisectors of a triangle passes from the same point, by giving an order to say mathematica to draw the bisectors and see if they passes from the same point!

i think research profiling will become a 'must have' in future phd. thesis manuscripts. most of the US scholars have already started adding a research profiling chapter before literature review in their thesis but it is still 'good to have' not a mandate in thesis. but seeing its utility and scope i think it will become an important part of thesis in future.