People learn in a variety of different ways yet lecture is the most common form of material dissemination. This is wonderful for the people who can soak in all of the information and draw conclusions themselves. This leaves many people behind if all they have time for is writing down facts and attempting to keep up with the basic material. Since most courses in high school and college no longer require intense critical thinking, a quick memorization of facts will allow most students to succeed and think they "KNOW" material. When asked to apply it many are unable to. Interactive engagement techniques do not require the removal of lecture from the learning process they just put less emphasis on it. Lecture is the ONLY way to present enough material in a college course and is critical to the active engagement techniques. Students must be given the basic knowledge before they can be left to begin their own critical thinking process.
We know from research that people learn by linking new concepts to concepts they already have a model for. Most of these models are incorrect when it comes to astronomical and physical phenomena. A student who has misconceptions may still think they understand the material and be able to respond correctly to some questions. However, when a question specifically calls out a known misconception, the model the student is using to reason through the question will lead them to the incorrect answer every time. What active engagement techniques employ is social conversation. Lecture tutorials are one form of this learner centered engagement. Students are given a 20 minute lecture on a topic such as the seasons. Then they spend 20 minutes with a partner working through a socratic dialog (in their lecture tutorial workbook made up of research validated questions and "fake" student responses). The pair works on coming to consensus and discussing the reasons for their answers on each question. As the students work through the dialog the concepts become more challenging and the misconceptions are challenged. Often students are required to look back at previous answers (known to be commonly incorrect) after some misconceptions have been challenged. Students are engaged in their own meta-cognition and are forced to confront their own and others ideas. This active form of discussing and defending your ideas allows for misconceptions to be overcome and new concepts to be better rooted in the brain.
For those of you who think this is useless. We performed a study of lecture tutorials in our classes. We split the classes into the top 50% of students and the bottom 50% of students. Before lecture the top students are scoring 50% on concepts not yet covered, those at the bottom are near 10%. After lecture BOTH groups are around 50-55%. This means lecture is helping students catch up with the basic information they may not have had. However lecture only got the class to FAILING! After a lecture tutorial in class, both groups are now performing at the 70% level. TWO WHOLE LETTER GRADES BETTER!!! This is why we say lecture is not the important part of the course because the student engagement is helping everyone.
So if lecture is only a means of giving out the information then there is not a critical need for professors to stand in front of the classroom at this time. We can hire actors which are far better at the job of dictating and making material exciting and record it. The professors job becomes important later when students have questions not for being the talking head.
Additionally, there is the bogus idea that a revised theory should still be considered a theory. Instead a revised theory is now no more than a hypothesis, requiring fresh predictions (to be tested against new observation not previous data) and verification and requiring the fresh application of Occam's Razor (since a revised theory is also usually going to have additional complexity to patch up the previous theory).
While I cannot completely disagree with your argument, theories are just that, a working hypothesis. We have a model that describes everything that happens in the universe around us. We call this model a theory. It does an amazing job predicting 99.999% of everything we see. Then someone makes a discovery that contradicts some of the assumptions and outcomes of the theory. We go back, look at the physics, and adjust/revise, the theory so it can explain all of the observations. Its not that the theory is broken or is wrong, it could not describe EVERYTHING. Every theory must explain existing data as a start and then make predictions about the data we will take in the future. A great way to test a theory is to see if it can model existing data. The data don't change because you changed your theory but if you need to make a new prediction then you need new data to test that piece. A good example of this is Newton's Laws of Gravity. NO ONE can argue that gravity exists and that Newton's Laws work for almost everything. Well they didn't work for Mercury's orbit. When they compared the predictions of Mercury's orbit from Newton's Laws, they found that the model was off by 43 arcseconds every centrury. Does that mean Newton's laws don't work here on Earth? No, it means we needed a new model of how gravity worked in more extreme situations. This is where Einstein's Theory of General Relativity comes in. It explained the precession right away and we have used Relativity since then to explain motions around black holes and other extreme objects. Newton's Laws still work perfectly fine with in the errors of measurement for everything else.