Shoot...as soon as I posted this I think I found an answer. It seems the input value is of the long data type (a 64-bit integer). I'm also reading somewhere online ( here) that Math.random() in Java uses a seed of the data type long, but actually only 48 bits are used for random number generation. I'm assuming that this is how Processing works too. Anyone know for sure?

http://processing.org/reference/randomSeed_.html

Processing's own version for randomSeed() uses an int data-type as its argument.
But I bet it internally calls the original Math.randomSeed() under the hood.

Anyways, if the original uses 48 bits and the wrapped 1 uses 32 bits, 16 bits are lost. 
But I guess 4 billion different values are enough for seed variation? 

OK, thanks!

Actually...maybe randomSeed() in Processing does take a long after all - when I purposefully make an error by putting in a float, the error message says the function expects a long.