Apparently I've been beckoned. X)

I'm sure we could all theorize endlessly on the relationship between SSRIs and dreams. There are so many variables at play.

First, there are the different phases of effects of SSRIs. You start with the therapeutic lag, the two-week build up to full effects. The reason it works this way is because when you first start on SSRIs your serotonin autoreceptors like 5-HT1A are sensitive to activation, and those autoreceptors actually reduce serotonin levels, thereby rendering the effects of serotonin reuptake inhibitors useless until these effects are overcome by receptor desensitization. This is also the reason that serotonin reuptake inhibitors are not recreational, unlike serotonin releasing agents like MDMA which flood the synapse with enough serotonin to overcome this inhibition immediately. However, just because serotonin is being depleted doesn't change the fact that these autoreceptors are being activated, and they have other effects of their own as well. Receptors like 5-HT1A and 5-HT2A actually increase dopaminergic effect, which makes sense as serotonin and dopamine often have inverse relationships, and 5-HT1A lowers glutamate and acetylcholine while 5-HT2A increases GABA activity and anandamide synthesis.

The dopaminergic area of the brain is necessary for dreaming, and increasing dopamine levels (as long as it's not enough to stimulate you and keep you awake) will enhance dreams. Lowered glutamate levels increase dissociation of the mind from body and upregulates dopamine D2 receptors, which also increases vividness and keeps you asleep better. Lowered acetylcholine drops lucidity a little bit, but probably not to a significant degree in this case. It's actually inverse correlated with dopamine here, so just kind of think of them as going hand in hand - you don't want too little acetylcholine just like you don't want too much dopamine, but a push in either direction is nice. GABAergic are known to increase dreams in small doses as well, and anandamide is an endocannabinoid, which means that it stimulates a release of phenethylamine and activates D2 as well. Too much of either of these will also inhibit dreams, but again, that's probably not significant in this case.

Once you've finally overcome the therapeutic lag, what I would imagine is going on is that these same receptors are being activated and so these chemical changes are still occurring, however serotonin is now remaining as a significantly active constituent. If I had to guess, I would say that this is where a good deal of the intense dreams come into play. Higher serotonin levels seem to increase vividness of details and definitely increase emotional intensity. Translate this into already enhanced dreams and what you've got is a dream with a stunning (sometimes literally) level of complex architecture and deep, meaningful plots. Furthermore, 5-HT2A is the psychedelic receptor.... Serotonin alone may not cause the level of colorful, chaotic visuals that something like LSD or psilocin would, but you'd be quite mistaken to think that it doesn't have lots of dissociative and ego-crushing potential. It doesn't surprise me to think that overstimulation of serotonin receptors (like with the higher SSRI doses) could cause bizarre, unexplainable, reality-bending dreams. However, I'd say there's another possibility for this, too.

SSRIs have been shown to actually decrease serotonin levels in a surprisingly short span of time. This is because serotonin levels aren't actually being increased by SSRIs at all, it's just being held in the synapse longer and therefor the receptors are being activated more strongly. The brain responds to this by creating less serotonin. This is one of the reasons SSRIs are actually a pretty bad idea for most people to take... and likely why they turn most people who've been on them for years into emotionless zombies. As I believe has been said, serotonin levels are lower during REM sleep. Personally, I think this mechanism is facilitated through either GABA-A or 5-HT2A (or both). That could in part be why dreams are enhanced in longer-term use of SSRIs. Dopamine levels will also be increased because of this, which as I said earlier will enhance dreams. Anything beyond that is really speculation for me (as if it all wasn't already), but I do have something more to add.

SSRI withdrawals are actually a result of very high serotonin levels. Ridiculous, isn't it? But it's because quitting SSRIs after that phase where serotonin has been downregulated forces the brain to overcompensate to try to restore serotonergic activity to normal. Unlike the normal healthy amounts of lowered acetylcholine seen in dreams, this overstimulation of serotonin receptors will cause a low enough level to significantly reduce lucidity and vividness of dreams. In addition, anticholinergics (i.e., drugs that work to directly inhibit acetylcholine) are known to frequently cause nightmares. So that's probably where that connection comes in.

My two cents. I'm kind of tired right now, so sorry if it seems like I rambled a bit.