Is liking personality descriptions with no supporting evidence as valid as liking personality descriptions that are supported with scientific evidence?

There is a discussion over here about the value of the Myers-Briggs tests. One person said there is no evidence to say that the M-B tests tell you anything useful. Others argued that they, personally, got something useful out of the test.

It made me wonder whether there is any difference in the value of information to people when comparing information that is only valuable to the individual for personal reasons to information that has scientific evidence to support it.

For example, a lot of people find astrology or tarot readings to be valuable and accurate. Yet there is no scientific evidence that these readings are any more accurate than random characterizations of people.

It’s easy for science types to say these people are idiots because they make meaning out of the noise. I’m trying to take the value judgments out of the equation to ask whether, all things considered, it matters if personality descriptions are scientifically valid or only personally valid. Is there any significant adverse impact if one finds a random personality description to be helpful for ones personal growth when compared to a scientifically validated personality description?