The competent bacteria I use have a transformation efficiency of >10^9 colonies/ug. So officially 0.1 pg in a transformation should give 100 colonies. So, in the real world ng's or pg's of intact DNA should give colonies, and it can be a good idea to plate a dilution if you predict it might be too crowded. Other than that it's just the practical matter of using a volume that won't be entirely absorbed by the paper. It's not rocket science, if you use too much DNA and it's a clone (not a library) it's not a big problem if more than one plasmid goes into a cell.
PS.

I'm one of those stingy people. Remember with global warming, limited resources, recession, etc, why waste resources? I pipet into a circle marked on the paper so only a small volume is necessary. If I sent more than 1 ug it would be a waste to put it on paper when in a tube it could be used directly for a cloning.