Hmmm. I didn’t realize that this was such a hot issue for you. I never really knew what Alex Williams was referring to until I started reading Werner Gitt’s “In the Beginning was Information.” His thesis does not exclude Shannon’s mathematical theory of information. Rather, it merely stipulates that Shannon’s theory occupies the lowest level of information, namely statistics (as opposed to the highest level, meaning).
Jeepers, I don't know where Drs. Williams and Gitt got their idea of what Shannon information theory actually is.
Crucial to information theory is the separation of the content of the message from its conduit. That is, it draws a clear distinction between the "message" and the "medium" of its conveyance. IOW, pace Marshall McLuhen, the medium is NOT the message.
Shannon theory deals only with the medium, the conduit. The conduit is mathematically described; it can carry all sorts of messages whatsoever. But you don't have "information" unless a message is successfully received by an intended receiver, and that's why the structure of the conduit is so important.
In short, if I had to say where Shannon information theory fits into Dr. Williams' autopoetic model (and it seems it must, for the Shannon model is universal; i.e., applicable to all communications), I'd simply describe it as the means whereby the "message" of the inversely-causal metainformation is timely conveyed to the various levels of the autopoetic hierarchy. Shannon information theory is totally indifferent to the actual content of the message. Again, it is the medium of its conveyance, not the message itself.
The actual source of the message is not to be found within the Shannon model. It seems to me that Williams' inversely-causal metainformation may well have an extra-mundane source. Whatever. If it is to be successfully communicated, it must go "the Shannon route."
Thank you, dearest sister in Christ, for pinging me to this interesting discussion with GGG!