Actors as a special case of dataflow

The theme of actors in recent years is gaining popularity and some authors are presented as a hit
season, although the concept goes back to 70 years. Erlangians are especially zealous - alone
the headlines are worth it - At the forefront of virtual machine design , the headlines of Soviet newspapers are remembered.

The dataflow principle is my favorite programming principle. It is obvious, universal, and at the same time it is difficult to use it directly both for computers and for the brains of programmers. In a nutshell, an operation is performed when all its operands are ready, and its result is an operand for other operations. From this point of view, the actor is a simple dataflow operation:
.
where In is the input messages, Out is the weekend, State is the internal state of the actor. Therefore, only basic dataflow constructs can be implemented, and the actors can be defined at the user level. However, there are many actor libraries, and very little dataflow. In other words, the authors of actor libraries do not want to give access to the inevitably existing dataflow inside. Is this due to distrust of the user or misunderstanding that they "speak prose"?