I am not in academics but was for some time, and I agree with some, but not all your sentiments.
The argument of academic departments making or losing money with trainees is nuanced. I don't believe most departments approach this as a financial win/lose as you describe in the decision to have trainees or how many. Academic departments have a mission to have trainees. Academic pathologists in part take jobs there to train trainees. It is not as simple as a calculated "how can we make more money" decision and describing it as such does not ring true to me.
the argument that pathology departments "lose" money with trainees is true from a certain perspective- one that is aligned with your description of how you think the department operates- namely, in an effort to make money for the university (rather than fulfilling research and clinical objectives for the university). If there were no trainees, they could argue according to your model, their staff could focus on signing out cases rather than providing lectures to residents and students, mentoring residents, and spending time reviewing each case with a resident to teach them something. They would not have to correct residents' bad diagnosis or reports. They could effectively spend their entirety of their FTEs signing out cases (basically like private practice). With this in mind, the department, assuming a stable amount of volume, could hire fewer pathologists to do the work. Each pathologist would certainly, in this scenario, generate far more revenue for the department at lower cost than what they can do with trainees. Having trainees from this perspective COSTS the department money as trainees cannot perform any billable duties. The amount of money the department gets from CME for a resident does not cover the amount of POTENTIAL revenue lost by having to deal with them; I think this is even true when you substitute trainees with PAs.
However, this scenario above is and can only be hypothetical. Academia is NOT private practice; their mission is to do research and to train residents. So to assume both that they make decisions to hire trainees or not hire them for purely economic ones is flawed. Academia will have trainees. They will leverage their use for their mission AND for economic reasons, and there is no one approach that will be true for all academic departments. From my experience, the rationale to hire more trainees is far more practical and a result of departmental budgets than anything to do with raising revenue for the department or university from CME. I actually believe that departments do not get any tangible economic benefit from trainees; but that does not mean that they do not actively pursue them for entirely selfless reasons. The decision to have or increase the number of residents is proportional to the amount of pathology services being rendered. The consideration may be whether a PA or a trainee should be brought in (or less likely, additional staff). I think far more relevant to the decision making (over how much revenue will come from CME) is where the money will come from. Hiring another PA means having enough money in the departmental budget to be able to afford it. This budget is tightly controlled. This means likely cutting costs in the staff budget or asking for more money from the university or being allowed to use more of your revenue for this purpose- revenue that other departments and research initiatives are dependent on. There is no guarantee that it will be allowed, and even if enough money is available within control of the Chair, the decision must be to prioritize these funds for a PA when there are many other projects within scope of the department's responsibilities that will compete for those funds. However, another trainee usually comes from another bucket of resources entirely. So even though the gestalt of having a trainee may cost the department (or really, the university) money, from a practical perspective, this is seen as free labor to the department because it comes from another bucket entirely- the CME office. Some departments will consider only this when they apply for positions- substituting trainees for PAs- while others (the good ones) will consider more trainees when there is more volume than their current residents can handle and they determine more trainees can be trained at their facility with proper oversight and training. Notice that while subsequent employment opportunities are a consideration and very important to the department for their trainees- the department itself has NO oversight over the total number of positions available or the larger marketplace. All they know is they could train one more.
It is understood that some departments make money for the university, while others lose money. The university will take money from the departments that make money to make up for those that lose money. Pathology, in general, makes money for the university. The departments that lose money- like endocrinology- still have these same issues, If the department and university ran on the principles many of you ascribe to academia, these departments would not exist.