During the run-up to the 2008-2009 subprime crisis, there were housing brokers who relied heavily on identitarianism and trust to market adjustable-rate mortgages to various groups. Matching up sellers to buyers by race/religion/gender etc. was a pretty effective technique for getting signatures on loan agreements. The result was many trusting people ended up with loans that blew up on them a few years later, resulting in many loan defaults and resulting economic collapses.
On the other hand, using such tactics is helpful in getting accurate information to people. If you ran a public health campaign aimed at reducing infectious disease transmission, a positive goal by almost any measure, then matching the message to the group would likely improve adoption of various effective measures (handwashing & general sanitation, for example).
In general, though, people are better off if they can extract the information from a sales pitch and make their decisions on the basis of rational analysis, not on emotional resonance.
The information can mean that the other does in fact understand you better, even if they don't know you personally.
I would argue that it's a valid heuristic and not out of place in the modern world.
You see it all the time in social interactions where individuals want to be relatable. Of course I can be manipulated, but that doesn't mean it isn't of value
Once a person knows more (relevant, accurate & truthful) information about another person then they will have a better mental-model of them, and I think that this will often make them feel like the other person could know them in better detail in-response as a person (even though there has only been one direction of information-flow so far).
The error could be huge if the base-assumption is wrong, but if it's not wrong then we could already know some specific details of the other person's mind quite well indeed (meme-complex-detected!).
In certain cases, especially around the description of qualities such as approachability and humility, hearing of these things from a speaker and in particular if they are common with the listener, could reasonably suggest that the speaker might also recognize these same qualities in the listener, and this could imply a possible connection without any bi-directional information transfer needed.