Jar,
I found an article on pubmed that shows (as you suggested) that coopting previously-existing phenotypes is a realistic process for the evolution of these complex systems. It also goes some way to answering the original question on evolution of hormone receptor systems.
Here's the abstract of the article:
Immune function requires intercellular communication. The vocabulary includes messenger molecules closely linked to the immune system as well as more widely acting messengers such as hormones and neuroactive substances. To try to bring these together, we have used an evolutionary approach. Materials that resemble hormonal peptides and neuropeptides, previously thought to be restricted to multicellular animals, are present in protozoa, bacteria, and higher plants. There is also evidence for substances in microbes that bind hormones and other messengers, which resemble receptors of vertebrates. Therefore, we suggest that the molecules of intercellular communication probably arose much earlier in evolution than the endocrine, nervous, and immune systems. This insight provides new understanding of messenger systems in vertebrates, as applied to the immune system, as well as new insights into possible disease mechanisms, including those that involve autoimmunity.
The article is available at
http://www.jimmunol.org/cgi/reprint/135/2/816S