Really it depends on many things and it needs to be a collective decision and sometimes it's trial and error, but you should never rule out the operators opinion if he's run alot of metallics.
You mentioned "overprinting" as if the solid metallic image isn't knocked out, that's usually a problem and if you can't fix it in prep it can be alot of fun on press. How big of a problem depends on what paper you're using (you'll get away with alot more on uncoated paper than anything matte or gloss), whether it's a light screen or solid image that's printing over or under it, run length and the consumables.
The company I work for knows to avoid surprinting a metallic if possible, they will usually knockout and trap the metallic.
If not surprinting, I will run the metallic last as metallics tend to sensitize the plates that run after it. On coated paper, the metallic will be pulled off the web onto the blankets of the later units, since there's actual metal in the ink it will damage the later plates and they will begin to print wherever the metallic is, offset paper isn't as much of a concern. This depends on run length and the actual inks, some times it's not a problem, but all things being equal, I run it last.
In the unfortunate event that it is surprinting on coated paper and we can't do anything to the customers file, then you have to get creative. As Gordo suggested the print rotation for process I use is KCMY, this is common for opacity but ink tack becomes a concern. Tack should run high to low which follow opacity. I think Gordo touched on this with the "receiving" silver, I don't know that term, but I assume it's a higher tack ink.
I ran a job from hell once, A large gold solid image knocked out for 4c lettering but surprinting black lettering, MICR black and gloss varnish. It was really unique because normally I have to run MICR last due to blanket take-off, but varnish had to run last. I makereadied the job Gold-K-C-M-Y-MICR-Varnish which was the most logical. The gold wouldn't print under standard tack K ink and Prep wouldn't knock out the K trap for the gold (other than to cutback screens). I switched to K-gold-C-M-Y-MICR. Then the Gold was eating up process plates within 30k imps (1,000,000 imp job) and the manager said "you guys will just have to fight threw it". The next shift came in and switched rotation once and couldn't run much of anything. I came in the next night and threw the book out the window. I think I wound up running Hi tackY-K-C- Low tackM-871 gold-MICR-Varnish. I had the K and M screens cutback and pushed the K density through the roof . The idea was to run the K density high enough so that the gold wouldn't even print over it and cutback certain screens to make up for my wacked out rotation and high K density.I had several extra MICR plates made, and just changed them every few hours, changing 1 plate on a Harris is no big deal but 10 takes almost an hour and gets expensive. It worked and didn't look half as bad as the previous tries, but I haven't seen any jobs like that on press since.