Our understandings of each other at any point in time can, at best, be visualized as Venn diagrams. At worst, there is no overlap, no comprehension, no mutual understanding.
Each of us uses language in unique ways and each of us understands the semiotics that underlie speech differently. This is why we can never expect another person to entirely understand what we mean by any statement, even when we know that person very well and even when we are certain they are very well disposed toward us.
It is a fundamental characteristic of language, speech, semiotics, human biographies, and the human brain that two people will never be able to make perfectly overlapping Venn diagrams concerning any utterance spoken between them, no matter how trivial. You might guess correctly once in a while, especially if your context is very limited, but as a rule you can never be sure of your guess and you will very often be completely wrong, especially if the context is complex.
Interpersonal contexts are virtually always complex. FIML practice is designed to work in real time with complex interpersonal contexts. FIML helps partners understand where their Venn sets overlap and where they do not.
If you do not frequently pursue in real time where your Venn sets overlap and where they do not, you will have a bad time. It cannot be otherwise because the divergence in mutual understanding that accrues between many poorly understood Venn sets will snowball.
I can’t think of another way to pursue the Venn sets of interpersonal communication besides FIML practice or something very similar. If pursuit of these sets is left to “take care of itself” or done solely or mostly with extrinsic generalities, it can’t work. It can’t work because generalities are the enemy of interpersonal set analysis, while things “taking care of themselves” is the mother of all generalities.