Coral have sophisticated endotoxic recognition and signaling capabilities.
Fish do not.
Humans do.
The original article referenced here (and thread itself) contains the links that document the above: http://www.freerepublic.com/focus/f-news/1823622/posts
You are assuming facts not in evidence. The authors of the actual paper did not engage in such flights of fancy. They determined an ancestral Toll-like gene was present. Similar Toll-like genes are found in fish. These Toll-like genes and more are found in humans.
Only one Toll-like gene is involved in endotoxin recognition in humans. This is TLR-4. It is found in most ray-finned fish (deleted in at least one species), as I demonstrated on the previous thread in which I tried to pound this information into your head. TLR-4 is involved in endotoxin recognition in mammals but may have a different function in fish. It is not present at all in coral.
Unfortunately your brain appears to be refusing new synaptic connections right now so the information slipped away as if down a clear mountain stream without making a dent in your memory last time I presented it. Maybe I'll have better luck this time.
Fish do not.
Humans do.
How can you say that "fish do not [have the relevant genes]" if no lobe-fin has ever been sequenced?
To make a claim about fish in general you would, at the very least, have to check ray-fins, lobe-fins, hagfish, cartilage fish, bony fish, etc etc.