It depends on what you mean by best. The intent of PPLNS is to discourage pool hopping and
reward long term mining. Whether it accomplisehes that is hard to say.
What I don't like about PPLNS is that they select which shares will be counted based on some algorithm.
Anyone who knows the details of the algorithm, specifically the value of N, can time their mining
sessions accordingly to gain an advantage.
PPS decouples the payout from the mining and pays a set amount for each share, like Nicehash does,
Prop is pure pooled mining where everyone gets a their fair share based on blocks found.
That's not true, about PPLNS - that exploit is exactly why PPLNS exists.
PPLNS was created to close an exploit in proportional-style payouts, where you could join the pool which solved the most recent block (and as such, their share counters reset.) The reason this works is because *if* the pool gets another block quickly, the "new" miner who just joined will receive an unfair portion of the payouts compared to those who had been mining for a longer period of time. If the pool did not find another block "quickly" (as defined by your pool hopping algorithm), then you would switch to the next pool which found the most recent block, and repeat.
The reason this doesn't work with PPLNS, even if you know the value of N, is because you cannot force a payout "round" to be short. The last N shares are unaffected by any found blocks, therefore no counters are reset. It is, in effect, a rolling window of time. No matter when you join/leave, you're not going to gain an unfair advantage.