User blog comment:Alejandro Magno/Ideas for ExE/@comment-25371339-20141012235536/@comment-5982810-20141013011250

Because this idea has already been proposed/considered at least 3 times before you brought it up. I considered it already before I even published E^, and it was independently proposed by Hollom and later by Vel. So far no one has done more with the idea than merely propose it as a possible extension.

One obvious problem at the outset, and already identified by Hollom, is that since ExE uses hyperion expressions as delimiters, it leads to confusion if it's arguments are also hyperion expressions. This is one of the reasons I opted not to go this route. Granted this problem is easily remedied in a variety of ways, but even then all the work is ahead of you. You still have to define the delimiter set, determine a proper ordering, assign fundamental sequences accordingly, and simplify this into a concise set of transformation rules.