Academia
I find them extremely useful, they are a launching point like Wikipedia. I am currently writing one and this is how I was taught to approach them.
Haha, well it isn't that simple. I use them more for the humanities to condense complex topics down to major principles. We also use them as teaching tools, for people that likely won't continue in academia. The big problem with these tends to be in the harder sciences, unfortunately. There are some journals which are extremely trustworthy for them due to the process of submissions, like Annual Reviews. These ones tend to be invite only though. Lots of people use systematic reviews as lit reviews too to launch off their own work and identify gaps. But you are not wrong; the issue is with junk science, and I would also like to soapbox here about the lack of education, even in graduate schools, on how to identify journal trustworthiness & prestige.