http://ccrc.tc.columbia.edu/media/k2/attachments/effectiveness-online-response-meta-analysis.pdf
https://www.dol.gov/odep/documents/SoftSkillsMastery.pdf
Boulderjohn - here are 2 studies to contradict you. Also my FAA Certified Flight Instructor certificate and renewals indicates opposition to the comment "I can assure you that the very worst way to have people learn content is through instructor lecture". Blended is the best way to instruct. Since I renewed my CFI in October, the FAA still has not changed that. Nor do these studies. Online can be useful but you also need face time.
First of all, nothing Boulderjohn wrote led me to believe that he was advocating for online-only scuba training. I understood him to say that certain information is best conveyed in some form other than lecture:
...I can assure you that the very worst way to have people learn content is through instructor lecture. Both the process of having students read the materials in a book and then go over it with an instructor and the online options impart that information much more effectively than do lectures.
Besides, Packrat, the findings put forth by Columbia University's Teacher's College are much more nuanced than you are suggesting. The report you provided a link to was in response to a meta-analysis conducted by the Department of Education. I haven't read that DoEd report, but based on my experience as a social scientist who researches and evaluates government programs, including training and education programs, I imagine that the study team reviewed studies, evaluations, and other literature on the effectiveness of college-level online courses. That available literature suggests that, in many cases, college students learn more in online courses than in face-to-face settings. According to TC, the literature available on fully online, semester-length college courses does not fully support this finding. I didn't look to see how "fully online" is defined, though I would think it involves no in-person contact with the instructor or instructor's assistants.
Secondly, the Department of Labor report is about imparting job relevant soft skills (e.g., interviewing, team work, communication, etc.) to youth and incorporating technology into the training curriculum for better outcomes, rather than relying on face-to-face training alone.
So, you offered two reports that concern non-conclusive student outcomes in semester-length fully online courses for college students and increasing the success of teaching soft skills to high school and college students through the inclusion of technology in the curriculum. I could be wrong, but I believe you're cherry-picking studies that appear in the top 5 search results on Google to support your argument. An argument, it appears to me, that is based on your misunderstanding Boulderjohn's posts. By the way, your actions only reinforce your criticism of online-only training:
Pop onto the internet to get a quick answer with no understanding.
Anyway, I don't believe there is a "best" way to do anything. There is current wisdom about what is "best," and then that changes as better evidence becomes available and a new "best" is defined. In general, though, in developing and delivering any training curricula, you will need to determine the audience, the information needs, the expected outcomes, and the resources available. I imagine the development of the curriculum for the FAA's Certified Flight Instructor certificate program went through this same process. What does that curriculum look like? Online modules, some in-class time, and then taking a plane out to show you know how to fly it?