New MatrixOp function?

I would like to see a new function added to the MatrixOp operation, namely something like 'LargerVal(a,b)' which returned the larger of the two arguments. This would augment the existing MatrixOP 'greater(a,b)' function, but instead of returning an unsigned byte (logical) value, it would return the larger argument value. If either 'a' or 'b' were complex, I would suggest use of magnitudes for comparison, but returning the data value and type of the larger. Even if the new function were restricted to real arguments, it would be useful. I know such a comparison could be done with a multiple-step sequence, but to use MatrixOp's fastest speed capability, I would hope a single function could be implemented.