interesting description. I see that the entry symbol map treats something what can be interpreted as a very special case, namely of a Clifford algebra. On the other hand, when one has a filtered algebra whose associated graded is commutative, one can define a symbol map in much more generality, and the notion can be sheafified. The result is far more general than both notions (symbol map for Clifford algebras and symbol map for algebras and sheaves of differential operators) and widely used. Once I have a thought I will try to write something along those lines (or if somebody else is quicker than somebody else will).

]]>wrote something at *symbol of a differential operator*