Arnold (2005) is the first mention (I think) of using transits as a potential form of communication. Arnold suggests that advanced civilizations could embark on ridiculous engineering adventures and launch one or multiple crafts into orbit as a way to long-term communicate their existence. A civilization could launch a single, large object that would block out a substantial amount of the star (like a Dyson sphere), or they could launch objects far from circular, whose transit curves would have different ingresses and/or egresses.
Personally, I doubt this would ever happen with the intent of communication. Such a project would take so long to complete, and so many resources, that I doubt a civilization would bother. If they were to bother, I don’t think the intent would be as a form of communication.
Arnold mentions that civilizations could group multiple objects into prime numbers (his example is 11 objects with 1,2,3, then 5 objects). This just seems like pure fiction to me. I haven’t done any math or simulations, but I’m skeptical that it is possible to keep this many objects gravitationally stable while still maintaining transit alignment. And even if this could be done, it would quickly go unstable, with orbits deteriorating.
If people search trasist data for megastructures, which they have and will hopefully continue to do, I feel that they should just look for odd or anomalous transits, possibly even swarms. But I don’t think people should bother with looking for messages in transits.