slip0010 derivation retry fix

[evallozhe]

https://github.com/satoshilabs/slips/blob/master/slip-0010.md#test-derivation-retry-for-nist256p1

[richardkiss]

This is a great catch. However, I can't simply start using this new algorithm because legacy users may depend on the old broken behaviour to unlock their UTXOs. I need to think about this, but I may need to continue to do the derivation both ways and call out when it differs.