@gjhiggins:
Regarding the checkpoint/51%-attack topic: I think it's good that we got rid of the live checkpoints. I was (long time ago) an active member of the Peercoin community until I quit because development stalled and many problems (among them, just the checkpointing "feature") were not solved. They still haven't removed the sync checkpoints - their current development version at least makes them optional to be enforced.
There was some research in the Nxt community and they have adopted a "reorg prohibition" for 720 blocks (also called "rolling checkpoints"), although this does not prevent short-range attacks to exchanges as these normally require less than 50 confirmations for deposits. (
see here for some research on PoS currencies and the "Nothing at stake" problem tat can lead to reorg attacks with low "staking").
The easiest solution seems to me that exchanges that deal with small coins could prohibit withdrawals in other cryptocurrencies for a certain time period (e.g. half a day) when the origin of the money comes from a small and weak altcoin, or - even simpler - require more confirmations (100+) for deposits. The longer this timeframe the more difficult the short range attack becomes. In combination with a "reorg prohibition" like in Nxt that could lead to satisfying security.
I don't know if such an "reorg prohibition"/"rolling checkpoint" is easy to implement in Slimcoin, but even if it does not prevent short range attacks it seems to me to be a relatively good solution to protect from long range attacks as it does not rely on a single node.
For the "hard coded checkpoints problem": As far as I know, Bitcoin releases a "hard coded" checkpoint every time a software release is ready, and "the developers" of that release are in charge. However, they release a new version every couple of months, so that may not be entirely comparable to Slimcoin's current situation. We could designate a person for a timeframe (e.g. every 6 months) to do that (e.g. someone being responsible for the 2 or 3 next checkpoints). I for myself could imagine to take this task, although I would have to learn how exactly to do that but it may not be impossible to take responsibility for that task, because as far as I know there are not really C++ coding skills necessary (correct me please). (More difficult for me would be to compile an actual binary for all platforms with the checkpoints included.)