in my experience magnets are pretty reasonable methods and here's my reasoning:
1. For most applications we use Li-Ion ells for we can use protected cells without a problem, if the magnet shorts the cell the protection should prevent problems. So unless you are using unprotected cells, magnets are a reasonable solution with relatively low risk.
2. Most applications where you will be using a magnet are situations where the positive contact of the cell is RECESSED below the level of the shrink-wrap, so think about it, if the magnet tries to shift sideways, it will run into the shrink-wrap and simply be off-center rather than completely off to the side... if your magnet is capable of shifting all the way to the side, then chances are you didn't need the magnet there in the first place.
I have personally made my own solder-blob button top on an 18650 without any problems, cell works flawlessly, and the protection still works. This is tricky as it requires that you get the iron very hot and only have the heat on the cell very temporarily, the method I used was to "drip" a blob onto the cell first, then reflow it very quickly to make it stick after it had cooled off, then sanded the top of the blob down so it's sticking up just enough for my needs. This is probably not the best way to do it but I suck at soldering. Overall soldering Li-Ion cells is considered a no-no.
