After re-reading the patent, I discovered that they use 500 mL of solvent for 16.2 g safrole. I have been using 1120 mL solvent for 100 g safrole. If I followed the procedure in the patent, I would use 36.3 g safrole with this amount of solvent. That number is very close to the yield of ketone that I am recovering, so it might be significant. Mabey the solvent volume is the whole problem. I'm about to try adding safrole and nitrite in 5 g (of safrole) quantities until the ppt. forms. This might give me an indication of how much safrole this rxn can handle.
Osmium - the ppt. was a powdery solid, not gooey or slimy at all. It had the consistancy of PdCl2 or CuCl2. I don't know the weight exactly, but it is between one and five grams. I'll try some of the stuff you suggested next time I run it.