Correlation or causation depends on the design of the study. When it comes to surveys, those would be correlational studies. When it comes to studying animal behavior, those would be causation.
Absolutely. Most of the studies I have seen discussed or come across in psychology have been correlation-based. While many people are good at saying they don't know for sure what the study means, most people looking at it interpret it to have meaning that fits with their preexisting biases.
Any study's results are only generalizable to the population from which the sample was derived. Thus if the sample was taken from a population of Ohio State university students, those results are only generalizable to that population.
Yes, hence the problem with conducting so many experiments on college students.
Your complaint is with the media and how they report the results no the study's principle investigator.
Not only them. You also see a lot of the same problems in psychology textbooks, for example, and among psychologists. Psychologists are not immune to the problems which plague non-psychologists looking at the research.
I have no problem with any study's principal investigator. I may have problems with their conclusions, but prefer to read a study before I critique it. A broad statement that I have seen certain problems in a field or two does not invalidate the work of any particular person, or even the field as a whole--it simply says that I have seen an issue that the field needs to work on. And it does, to some extent--while psychology is still bad at questioning some underlying tenets, it is much more focused on, for example, cross-cultural research than it was twenty years ago.