James Kent made signal theory. There are problems with signal theory although it is at least a start. For starters the role of 5-HT2a in the visual effects is questionable. Also ST is admittedly completely unable to explain the more complex types of visuals but it is an interesting start regarding some of the mild "open eye" visuals. It is known that many of the visuals result from activity at higher levels cortical circuits rather than early regions such as the retina, thalamus (LGN), or V1. Additionally it is likely that specific cortical regions are involved in each "class" or "form" of visual effects. Imaging studies are going to be required to sort this all out. Additionaly lesion studies will also be required to associate specific cortical regions with categories of visual effects.
Also there is evidence suggesting that receptors other than 5-HT2a, and other 5-HT receptors are responsible for the visuals (Wallach's endogenous hallucinogen article:
Erowid.org: Erowid Reference 7404 : Endogenous hallucinogens as ligands of the trace amine receptors: A possible role in sensory perception : Wallach JV). Several compounds with 5-HT2a activity that are considered hallucinogens lack visuals. In fact many compounds lack visuals but are still hallucinogens (5-Meo-DMT, DET, MIPT, AMT....) there are also 5-HT2a agonist that are not hallucinogenic at all. This is a fascinating area unfortunately little is known.
There is not much information on this. It is very fascinating and a group in germany has recently published a few articles using imaging studies during DMT and Ketamine experiences (type DMT and ketamine into google scholars and a few should come up). SWIM left his bag in the office with the articles but will post the titles later on. Science does not even know where normal waking visual experience comes form let a lone hallucinogen induced visuals. However hallucinogens are a unique tool to study the processes in the CNS involved in subjective experience.