If there are more classic pools, why there should not be more than one p2pool network? Why the hell use some 3rd party proxy if I can directly mine on 2nd or 3rd network without problems (with hashrate that results many hours to share on 1st network). Proxy is more like classic pool than p2p.
Fragmenting the hash power into separate p2pool networks is just a work around to p2pool's current inability to reduce variance for small miners. If people are mining vertcoin long-term then the variance shouldn't be a major concern since it all averages out and a small miner will receive their average expected earnings over a long enough period of time.
If we wanted to consider the other extreme, the smallest p2pool network that makes sense at a 248.953 difficulty would be around 15MH if I do the math right, that'd average 2 blocks per day. If a single GPU miner has 300K hash power, then that'd be about 50 miners per network. With each miner actually holding 2% of the pool hash power, there'd be no problem at all getting shares, each miner would have a high average number of shares so the variance would only be based on the pool finding blocks, not added to the miner finding shares.
I guess it'd be interesting if p2pool could self-organize itself into multiple block chains and miners are assigned to an appropriate chain in this load balancing sort of way. The author of p2pool is working on some changes to support smaller miners, but I'm not sure on any of the details yet.
But trying to manually organize the 950MH on the original p2pool network into as many as 63 separate networks to hit the 15MH target above would be crazy.
Someone less lazy than me could calculate what hash power the pool could have for a single GPU miner to still have a 95% or 99% chance of keeping at least 1 share in the share chain at all times over time. There is a cap that p2pool tries to keep single miners at less than 1.6% or so of the whole share chain, so each miner having 2% is rather high. 30MH would make single GPU miners be at 1% of the pool power, and increase block/day (at current diff) to 4 on average.
If vertcoin grows substantially you do eventually end up with BitCoin's problem. The whole BTC p2pool network is over 1.5 days per block on average. However since the pool hash rate is around 180TH, it makes it hard for smaller miners to find shares. Over time it all averages out, but many small miners just can't stomach the variance of when they get paid.