For anyone who didn't read the article I linked above, they did a test on various types of hooks and two methods of leaving the hook in. The first was leaving the line connected and the second was clipping the line close to the shank. The results, quoted from the article:
"After four months, 78 percent of the hooks were still imbedded. Cadmium coated hooks poisoned 20 percent of the fish, and production of these hooks has been stopped. Bronzed hooks were less likely (70%) to be retained than tin-cadmium (80%), nickel (83%), or stainless steel (100%) hooks.
In a second test, the line was clipped at the eye of the hook, as advised by most existing C&R guides. One-hundred percent of the stainless hooks were again retained, while 56 percent of tin, 76 percent of bronze, 84 percent of tin-cadmium, and 88 percent of nickel hooks remained. Fish mortality was greater when all line was trimmed. Foster theorized that the lengths of line hanging from a fish's mouth kept the hook-shank flat against the side of the esophagus and allowed food to pass. Without the line, food could move the hook and close the throat. "
Also, this was done in half-strength saltwater, not freshwater. I'd imagine the time to rust out would be significantly longer in freshwater.