Cost-effectiveness isn't an issue with this customer. They're already ok with spending many thousands on the materials to have a fountain in the pond...I dont know what the actual load is to try and make a judgement, but don't forget to consider transformer cost and losses in your analysis. You will have two transformers with load and no load losses potentially 24 hours a day (depending of course on when this runs and the control scheme if it's not continuous). Two transformers will themselves have non insignificant voltage drop. Sometimes a step up step down doesn't pay for itself. Also for a fixed load like a fountain, perhaps just using a boost transformer if the voltage is too low would be more cost effective.