That's true. I think what happens is that researchers sometimes believe they need enough material for their papers to be published. For instance, they prove mathematical properties of their models which obfuscate the message and aren't necessarily relevant.

Also the target audience are other specialists in the same narrow field.