I love open content and use MIT OCW and other materials to prepare and teach my courses. But I don't think virtual labs will ever compete with real labs. In reality, your magnetic field measurements are complicated by the NMR coil upstairs and you have to explain why. In reality, data point number 7 does not fit the line. Even with random number generators, virtual labs disconnect measurements from reality and are not a valid substitute.

I've taught physics labs both ways and the students are happier with real labs and the learning outcomes are much better with real labs.

Unfortunately, at most universities, you'd be right, professors do not interact with students and there is no "real" communication. But there is already an alternative. Small colleges (less than 5000 students) with no TA's encourage communication and collaboration between undergraduate students and professors. I'm thrilled to be working at one. By far the best part of my day is office hours, working with individual students to better understand class or the textbook.