I just saw that and read the pertaining bits of apo6. In particular
RFC 107: lvalue subs should receive the rvalue as an argument
This would make it hard to dynamically scope an attribute. You'd have to call the method twice--once to get the old value, and once to set the new value.
The essence of the lvalue problem is that you'd like to separate the identification of the object from its manipulation. Forcing the new value into the same argument list as arguments meant to identify the object is going to mess up all sorts of things like assignment operators and temporization.
Which, without some form of explaination of how it "is going to mess up all sorts of things" and why those things are considered more important than the ability to validate the assigned value, sounds an aweful lot like "because I said so" :(
Examine what is said, not who speaks.
Silence betokens consent.
Love the truth but pardon error.