Do you have any published data showing that DFT or/and MP2 are good in predicting organic cations (comparison between calculated and experimental values)?
Generally "bests" method, including quantum chemical ones do not exist! To the topic of your question most important is which method balances between accuracy and computational time and cost. This depends which properties you would like to compute (including how large is your system). You can pay attention to the attachment, showing thermodynamics in solution of organic anion using B3PW91/SDD. For cation s teh same but the charge is +1. Usually this theoretical level is for inorganics, so in your case you can replace the functional with for example B3LYP (it is become standard) and some of accurate basis sets such as Dunnunig"s ones. In spite of that B3PW91 can be used, too. In addition DFTs provide meaningful theoretical information about large scale properties.