[gay_genocide]

This is just free advice. Build a disk based index then compare with industry standard database storage engines under real world conditions. Compare your work to others in the field such as Tokutek's fractal tree. If a certain index is faster in one specific way most likely it is worse in others, such as creation time, resource requirements, etc.

[hemen]

Thanks for the advice. Complexity is not much different than bTree for creation of hTree. interestingly there is more than 95% probability that the index is going to be faster than the existing btree.