Not sure about the exact mechanism of NP training but I can tell you that PA training does not utilize algorithm style learning. My library of required textbooks includes Harrison's IM, CMDT, Merck Manual, Netter's, Tintinalli's, Bates...the EXACT same books Med schools use. In fact, Med students and PAs take the same classes together at some programs.
I'm not sure if we're agreeing or disagreeing or maybe a little of both. Just because you used textbooks doesn't mean you approach problems the same way a doc might. I mean, come on, you don't really claim to have read Harrisons or Tintinalli during PA school do you? It isn't uncommon at all for an emergency doc to read Tintinallis cover to cover or an internist to read Harrison's cover to cover during residency.
It isn't so much the algorithm based learning as the algorithm based practice that I see so much these days. (Although I've been told by PAs that their school was definitely "fast-tracked" in that when they asked "Why..." they were basically told "That's just the way it is" and the instructor moved on to the next subject.
I work frequently with half a dozen very good PAs and see algorithm based thinking all the time.
If we both see a 8 month old well-appearing girl with fever, a cough, and a pulse ox of 96% at sea level, this is the way the PA thinks about it:
Peds fever....blood cultures, CBC, urine cx, cxr. If WBC>15 or <5, give Rocephin, see again tomorrow here or in the pediatricians office.
Whereas the doc knows the literature behind which studies the patient needs and which ones the patient doesn't, so when a new paper comes out, he can evaluate it and determine how his practice ought to change. He also understands the reasoning, uncertainty, and limitations behind each of the tests and when they need to be done and when they don't. For instance, based on current literature, an 8 month old girl needs a UA/Cx even with a confirmed diagnosis of RSV but an 8 month old boy doesn't need a UA/Cx if he's circumcised, well-appearing, and has no hx of UTI. 5 months, they both get a UA. A febrile kiddo with an abnormal RR, pulse ox, abnormal chest exam, or respiratory symptoms needs a CXR, most kids don't. A well-appearing febrile child without significant PMH who's had his 2,4, and 6 month vaccinations doesn't need a blood culture. A PA may see that this is how the docs around him practice, and adopt their style, but he may not understand the reason why (the rate of false positives in this group is so much higher than the true positives due to the introduction of the Hib and Pneumococcal vaccines that we do more harm than good by testing and treating the positives.) More likely, he'll just order all the tests in my experience, "so he doesn't miss anything." 90% of the time, it's fine. They are subtleties, fine points, but at rare times, important distinctions.