The built-in Array Size function is fine, if you have a 1D array. But for arrays of more than one dimension, things get clunky.
Drop the attached .xnode on your block diagram. Wire an array to it, and play with the array's number of dimensions.
The Matrix Size will accept a 2D array and have the breakout between row and column sizes. I have rarely needed more than 2 dimensions in my arrays. But I do like where you are trying to take this.