1/ In the early days of Litecoin (2011), there were threats of 51% attacks. To keep Litecoin secure, I utilized checkpoints. Checkpointing is a centralizing feature as it relies on the developers deciding which is the right chain. But for the beginning of a coin, it made sense.
2/ It's trading off some centralization for network stability and reliability. Satoshi also used checkpoints in Bitcoin in the beginning. Though one can say that it is similar to the idea of UASF as users can decide whether or not to run or upgrade to the latest checkpoints.
3/ I have also considered preventing large re-orgs in the code but decided against it. There is a very good reason why I didn't do this in Litecoin and why Satoshi didn't do it for Bitcoin. Satoshi said it best in one of his emails. satoshi.nakamotoinstitute.org/emails/cryptog… 
4/ If you spin up a new node with the same consensus rules as the network, you need to come to the same conclusion as the rest of the network as to which is the right chain. The whole decentralized consensus is pretty much based on this simple fact to be true.
5/5 If this is not true, then it's not decentralized as you have to trust someone to tell you which is the right chain. The whole point of decentralization and sound money is that you don't have to trust anyone.
