I think I've figured out the problem. The herbalist event(healing plant in the XML files), like a few others has the second or third options give immediate effects as well as a certain probability of triggering additional events. For example, for the herbalist event, it gives you the option of triggering event_healing_plant_1, 2, and 3 in Civ4eventinfos.xml. 1 gives you happiness for 10 turns, 2 gives you one whip

and a probability of triggering event_healing_plant_4 and 5, whilel 3 gives you 2 whip

and a probability of triggering event_healing_plant_4 and 6. The immediate effects work, but when you pick 2 or 3, it doesn't trigger 4, 5, or 6 the way it's supposed to. So something's wrong with the <AdditionalEvents> thing not working. There aren't a lot of events that use <AdditionalEvents>, but the ones that do are all like the herbalist event in that options 2 and 3 are supposed to trigger additional events which don't trigger, which matches Frekk's observations on what events don't work.
That seems to be exactly what's going on.
The 'immediate effects' (gain or loss in happiness) are all contained in HEALING_PLANT 1, 2, or 3. The health effects and population loss are all covered in HEALING_PLANT 4, 5, and 6 which are (depending on random outcomes) triggered by 2 and 3. It's 4, 5, and 6 that seem to be broken, 1, 2, and 3 all seem to work fine.
Just for reference, I had a look at some of the other events that use additional events. They all use the iEventChance tag (seemingly used only as part of an additional event) and none of them seem to be 100%. Most of them seem to be related to events where you have several options/choices. I recall from playing vanilla that these events print the chances for the additional event, and its effect, so if someone's airliner downs in your territory, and you choose to "claim the wreckage", you should see some effects plus a "50% chance for" some other effects. Same with a few other events besides healing plant, including defecting agent (one of our agents has defected to X), the airliner crash (one of X's airliners has crashed in our territory), and the fugitive event (you discover a fugitive from X hiding in your empire).
In the herbalist event it's easy to spot the failure because the choices don't seem to make any sense (why would anyone choose to have angry faces for no benefit, when they could get happy faces instead?). But the additional event effects are easy to go unnoticed in other events, because usually, there will be effects that do work, besides the additional events. But whenever the defecting agent, herbalist, horticulture, airliner, or fugitive events occur, at least one of the choices ought to say "50% (or 30%, or 90% depending on the event) chance for ..."
Some of these events, like the fugitive and the defecting airliner, are quite common. And I can't recall ever seeing the % chance for other effects mentioned in any of them while playing RevDCM, though I know I've had defecting agent and fugitive.
Everyone should keep their eye out for these particular events (herbalist, fugitive, defecting agent, horticulture, airliner) and see if there is ever a % chance for other effects mentioned. If not, then <AdditionalEvents> (or maybe <iEventChance>) is broken.