From an operational POV spend outputs are rudimentary after 20 blocks and could be removed from the index, just as zero-balance public keys. Performing a rich-list scan shows that 90% of the public keys is empty (fully spend). This is inherent of the Bitcoin Protocol but allows for a tremendous amount of pruning both of the transaction index as the public key index.
A blockchain that is 10% of its current size seems to be feasable and does only need to be accompanied by a (non-indexed) full transaction-log and a (non-indexed) block-log to be completely rebuildable.
From an operational POV spend outputs are rudimentary after 20 blocks and could be removed from the index, just as zero-balance public keys. Performing a rich-list scan shows that 90% of the public keys is empty (fully spend). This is inherent of the Bitcoin Protocol but allows for a tremendous amount of pruning both of the transaction index as the public key index.
A blockchain that is 10% of its current size seems to be feasable and does only need to be accompanied by a (non-indexed) full transaction-log and a (non-indexed) block-log to be completely rebuildable.