There is obviously a difference in the characteristics of runways with a hard contaminant and a fluid contaminant. The former being typically ice or compacted snow and the latter being slush or standing water.
With hard contaminant the main problem is stopping distance, whilst with fluid contaminant both acceleration and stopping distances are increased.
As has been pointed out you are allowed to derate which increases controllability in the engine failure scenario and may permit higher take off weights. It depends of course whether you see engine failure as the principal risk. If you think that all the other possible reasons to reject a take off outweigh the very rare engine failures on the runway (except of course in the sim!) then going full thrust would seem to give the best safety margin by getting you airborne in the least possible distance and time.
However the first question is can you delay take off until the runway is cleared? After that comes the question how can you give yourself the best margin of safety. Clearly reducing weight is a good idea. In reality this means keeping extra fuel to a minimum, of course absolutely no tankering. Using the longest runway is also surprisingly overlooked on snowy days e.g. the number of people using 26L in Berlin Tegel when 26R is right next to it and considerably longer. Everyone knows this stuff, but compliance is at times patchy at least when watching from the holding point and observing where people are rotating along the runway.