http://lmrk.org/corexit_9500_uscueg.539287.pdf The part people are focusing on are the "distillates, petroleum, hydrotreated light". That's the part they're calling "deodorized kerosene". What it is, is basically light aliphatic fraction of oil (or what's now a light fraction after they cracked the heavier stuff, etc.). That makes up 10-30 percent on a w/w basis. Five percent (or less) is propylene glycol, otherwise known to most people as the relatively nontoxic antifreeze. Another ten to thirty percent is made up by the one proprietary chemical in the mixture. That one is going to be a sulfonate detergent of some sort. That one is the only unknown in the mix. That's probably the part that should have toxicologists saying "hmm", though that shouldn't be a significant inhalation risk for the workers. The light fraction aliphatics (i.e. "kerosone") people are talking about is not going to cause any problems different from those caused by the volatile organics in the oil in the first place.
Now, most of the volatile organics are going to come off the oil spill relatively quickly). Once the slick is a decent distance from the origin, most of the volatiles will have been lost. So, if you were to treat the slick at that point with the corexit product (or treat the shorelines, etc., as well, obviously) then you would be adding a source of those light hydrocarbons that wouldn't be there anymore. That's true. However, the volatiles that will evaporate from this product will *also* be fairly quickly lost.
For the guys actually working in the cleanup, there may indeed be an exposure risk. Whether to the petroleum distillates in this product, or to the volatiles from the slick itself. Depending on where they're working. Much like someone using paint thinner in his or her basement, such things should only be undertaken by people with proper protective gear and knowledge of how to use it. That's not necessarily a problem with the product, though it may well be a problem with the worker protections being used with it.
EDIT: The other corexit product they're using (I haven't seen anything mentioning relative amounts of these) is corexit 9527a.
http://lmrk.org/master_ec9527a_msds.539295.pdf This one contains 2-butoxyethanol, an important cleaner in Windex and Simple Green. This would have some significant advantages, in terms of aquatic toxicology, in that it seems to be rapidly biodegraded. However, it can have toxic effects with large respiratory exposures. Once again, this might actually be a *good* product in an environmental sense. That may not mean it's completely safe to use it without proper protective equipment, or other means of managing worker exposures. I would certainly hope they're keeping pregnant women from being significantly exposed, given the possible teratogenic effects. If it's being applied entirely underwater, however, I really doubt there is much risk of significant exposure for most workers. Exposure risk would *probably* be limited mostly to those directly involved in handling and delivery. The stuff is very miscible in water, and being piped in deep under water.