Necessary? No, you can do whatever you want. But if you're making important decisions about how to live your life, wouldn't you prefer to make those decisions based on knowledge that's been validated as thoroughly as possible?

Science looks for statistical trends. But I'm just one guy. Something could fail for >90% of the population and still work for me. Conversely, something could work for >90% of the population and fail for me. If the costs of trying the solution aren't high, trial and error (guided by some basic scientific theory) might be more successful than scouring the journals for whatever self help theory happens to have a little bit of clinical support these days. A lot of these theories haven't been tested, and a lot of them that have weren't too rigorous.

trial and error with a sample size of one is great at convincing you of something that's not true. It's horrible at everything else.

The least rigorously tested theories in the softest of sciences are still orders of magnitude more likely to be right than something one person has done through trial and error experiments on themselves.

How could knowing if something works or not not be necessary?

But if something has no evidence presented then I don't understand why an educated person would care. We know that without evidence the advice/tips/pet theories of human beings are much more likely to be wrong than to be right.

Without any evidence I read these kinds of things like "here is some bullshit that is almost certainly wrong in a fundamental way because I am human and therefore have a multitude of biases that I am unaware of".

Doesn't mean it isn't interesting to read someones personal philosophy, especially if it's well argued and written. But why would you follow advice when you have no way of determining if it's good or bad advice? The more convincing something like this is to your "gut" the more likely it is to be a specific kind of bad advice.