When using mmpool it correctly shows in its dashobard web system its hashrate.
When mining on my p2pool though it shows a hashrate about half what the p2pool is actually getting from it according to p2pool itself.
So this problem seems to be cosmetic, somehow the Jupiter's web system is extracting wrong information from the cgminer instance it is using.
But it took me some time to finally migrate to having my p2pool as first choice pool because at first it did seem like it had real problems trying to use p2pool.
One thing miners stand to gain by using p2pool instead of centralised pools is control over which coins to merge.
Until recently no public pool I knew of merged I0C, GRP, CLC nor XGG.
A while ago mmpool reinstated I0Coin and at the same time also added GRP.
But as far as I know you still miss out on both CLC and XGG if you use any centralised pool instead of running your own p2pool node.
As more blockchains implement merged mining, people using centralised pools will likely miss out on more and more merged mined blockchains.
-MarkM-
thanks for sharing your experience, really appreciated.
How did you discover that your real hashrate was actually twice the hashrate reported by cgminer?
I have a more general question though: how is possible to improve the situation in a technical way? I've heard of other problems related to p2pool (latencies, flushwork). Are those problems real? Is there enough man-power, commitment and resources behind the project?