The point was really that space-filling curve indexes do not solve the underlying issues with polygon indexing, their oldness was not a factor. That type of indexing is so old that the patents (yes, they were patented by the big database companies) have long since expired. And yet they are rarely used commercially. There is good reason for that. There is a resurgence in their usage by people new to spatial indexing but it is pretty obvious that few implementors are aware of the extent of the computer science, which goes back into the 1960s.
The assumption of size was "large-scale", which in this day and age usually means massively distributed. A B-tree is not a solution to that problem domain.
The assumption of size was "large-scale", which in this day and age usually means massively distributed. A B-tree is not a solution to that problem domain.