All these different notions have their places. In the forensic world, generally speaking, both my training and my experience lends me to agree that less is usually better.
The autopsy report should consist of your anatomic findings. And that's really all, with the exception of the cause/manner opinion in some circumstances (ME's generally include this, and most pathologists who either also complete death certificates or advise coroners in completing death certificates seem to do so; non-forensic reports are another beast). But that doesn't preclude a "comment" or "opinion" section, laboratory results section, or a "clinical summary" or similar section (in the forensic setting this is usually usurped by an investigative summary written by a forensic investigator or similar non-pathologist).
In residency we did have a "clinical information" section in our autopsy reports, which consisted of a heavily consolidated version of a discharge summary in one or two short paragraphs. In residency we also had a "clinico-pathologic correlation" section at the end consisting of a summary of findings and why, taken as a whole, the cause of death conclusion was what it was. In forensic fellowship we had an "opinion" section, which was used somewhat variably by the supervising medical examiners -- some were paranoid about trying to anticipate and answer every possible lawyer-question and consequently tended to be verbose, while others did little more than re-write the bullet-point anatomic findings in a few short sentences. In retrospect, the opinion section helps me remember what I was thinking -- and in reviewing other people's cases, understanding why they concluded what they did -- but yes, a good lawyer may be able to let you paint yourself into a corner using them. Currently I very rarely add a "comment" section, usually reserved for undetermined cases, consisting of one or two sentences briefly explaining why it was undetermined.
I do think that in the training setting there is utility to writing a clinical information section and a clinico-pathologic correlation section, but you have to understand any autopsy (just like any medical record) could end up in court. So, in writing them, I would avoid being absolutist, avoid speculation, and feel free to reference relevant conclusions.
I don't have a good link to examples. My clinical information sections generally went something like: A 333 year old white male presented to ThisHospital's ER July 4, 2052 with shortness of breath and was admitted. His medical history was significant for coronary artery disease status-post CABG in 2044, chronic obstructive pulmonary disease on home O2 since 2048, deep venous thromboses, and nanobot nidus in the left lung following occupational exposure. His hospital course included magnet therapy, leech therapy, and hypothermia treatments. On July 7th, 2052 at 2330 hours he was found deceased on the bathroom floor.
But, everybody does it differently, and your bosses will likely have their own preferences.
As for all the different departmental specialists getting together to come to a conclusion about a patient, I always thought that was what morbidity & mortality conferences were for. Unfortunately they mostly seem to be within a department. We were rarely asked to comment on the autopsy findings, even preliminarily, at the M&M's which usually took place within a week of death (so usually before residency hospital autopsies were complete) -- which seemed to rather defeat the purpose. In the uncommon cases where the cause of death was complicated or unusual, I usually made/make a number of phone calls to those other specialists. Nevertheless, offering an opinion on the anatomic cause of death is still well within the training of an anatomic pathologist.