I never heard about the "10% rule" and I never applied it. Transferring TLC (silica on aluminium plates, Merck) results to preparative column chromatography on similar silica usually worked without changes, although the situation is a bit different in the column: the upper part of the TLC plate is rather dry when letting it run, while the silica in the column is not. Note that I was never a friend of the dry packing method and always prefered the slurry method to prepare the columns.
Anyway, I never thought about increasing the polarity.
Bad separation typically was a result of column overload, bad packing or cracking of the stationary phase, e.g., when the diethylether evaporates in the column.
+1 For citing the classic article on flash chromatography!
I can really recommend the technique and I used it ... more than once ;)