The collapsed "single-graph" is schizoid between E-nets and SARA. e.g: You cannot figure out whether to write the text of the resolution-procedure, or the transition-procedure, in a bubble. That is the consequence of an attempt to present the E-nets formalism, in SARA looks, with the kindergarten-grade plagiarism of copycat82.
The E-net formalism keeps the two types of procedures separate.
SARA represents the resolution-location functionality (or, the resolution-procedure) as a variable/setting (the "@output_arcs = ..."), within the SARA interpretation domain (associated with a node). i.e: With SARA, the resolution-procedure is within its transition-procedure - as a variable.
The cut-and-pasteful plagiarism of copycat82, keeps the two procedures separate (similar to E-nets), but with an attempt to paint/post them similar to SARA, i.e., as if there were a single-procedure per node. Therefore, in copycat82 figures, in its examples, the resproc versus transproc content, cannot co-exist. i.e: They are subcases of SARA examples, where either a decision-question is asked, or something is computed, but not both together.
In fact, a node in copycat82, necessarily, is a three-separate-entities, presented as if that were a single entity:
So far, it may sound as is you could concatenate them as a single entity, as a more-restricted version of what SARA does (i.e: restricting the output_arcs selection always to be the last statement), but the problems do not end there. The single example that we have of a figure, with its corresponding resproc, in copycat82, obviously, refers to the data-value before the trans-proc updates it. In other words, although copycat82 associates the res-proc with the exit-gate, the data-value it refers to, is the value at the entrance-time. This defies a simple concatenation to be a solution.
Guessing/inferring the exact path, it is plagiarized through: For ease of reference, Da80 refers to a single resolution-procedure - instead of two procedures of E-nets, as associated with a transition. At that point, it is trivial to merge-or-swap that resolution-procedure with (the interpretation-domain of) a SARA node. But copycat82 is lopsided. Instead of G to follow F, the triple-structure presumably expects F to follow G (as F is at the output side, and it reflects about the results of G, too). If we were to correct that problem of copycat82, with an assumption that F must precede G (but without any modification to data), then it is again a carbon-copy of Da80 (p.640) decomposition (w.r.t. the inner-steps, too).
In summary, copycat82 is aligned with Da80 (E-nets), although it is (attempted to be) visually-presented, imitative of SARA/UCLA. (Keep in mind that, copycat82 never mentions Da80 anywhere, in its text. The other prior art, including SARA, are mentioned only in the literature overview sections, but never credited as a source of any idea, in copycat82.)