it is used mainly for organzing large data in sorted order like table of a relaltion,e.g 1,00,000 employees table stored in a DATABASE.this table has lets say 10,000 blocks and finding for some data directly onto table on an averrgare requires 5000 block accesses.using B tree this can be reduced to log10,000 +1 block access in worst case for finding some record.now basic transfer unit is a block and block size is in range of 512 bytes.so using a BST will increase the height of tree and also will increase the space comeplexity,so more the branching less is the depth and in balaced search tree it takes o(height)=o(log 10,000) .
now with higher degree the branching factor increases and height decreases so to be more precise its o(log 10,000/log(ORDER OF B TREE)
but order of a tree is constant.so aympototcally it wont recude disk block accesses but practically it does.