Given Donald's comments about masq pins, I highly doubt he exhaustively tested anything more than two card combos. There likely aren't any broken two card mandarin combos; but it is quite likely that there are three card combos which neither Donald nor his play testers have every touched.
That being said, mandarin tends to be a weak card because it delays deck advancement by a turn (minimum) and slows down the cycle rate once you start playing it. So of the potential combos out there for mandarin, they will almost certainly need to be combos that can either use mandarin to increase cycling (e.g. library/mandarin/hamlet) or have good odds of getting a killer card earlier (e.g. mandarin/upgrade/goons). While it does give you 3 coin, as an opening, it cannot break into 7 or 8 coin territory, making it weaker than many other options (afterall 2 silvers give pretty good chances at hitting 6 and baron/silver is good odds of 7). Thus you pay a turn in lost tempo, are unlikely to get appreciably more expensive cards most of the time, and move slower later in the game.
There are times when none of these apply. In chapel deck, mandarin can be good terminal that moves the dead chapel or green and gives you a good payout, generally, though I want to chapel first and acquire the mandarin later - chapel/mandarin is a collision risk and worse, makes you even more vulnerable to attacks. Late game you might draw a silver and a gold and know that your odds of winning are better if you buy the mandarin to put back the treasure & hope to hit the last province rather than painfully draw through a treasure deficient draw deck.
For openings, I can really only see potential for mandarin if you are going to use it to get some high power hands with it that you otherwise would miss. Apprentice mandarin trades some poor early hands for some extremely strong mid-game hands. I think you might see a similar thing with mandarin/upgrade/goons or mandarin/hamlet/lib (particularly if mandarin is the only virtual coin out there).