Would it make sense to use a Prolly Tree’s root hash as the ETag for a collection resource in an HTTP API? Given an HTTP API that: - includes a collection res

Would it make sense to use a Prolly Tree’s root hash as the ETag for a collection resource in an HTTP API? Given an HTTP API that: - includes a collection resource to which items maybe added and removed - and for which supporting an ETag response that reflects the current set of items is needed - where it is prohibitively expensive to enumerate all items to recalculate the ETag It seems an incrementally maintained Prolly Tree would be an efficient mechanism and the root hash could be used as the ETag since the order of operations isn’t important. Do I have that right?