Effective stake calculation is wrong for node status and rewards commands

[torfbolt]

The current implementation shows an effective stake limit increase for minipools that are not active on the beacon chain yet. This is not in line with the rewards tree calculation spec (https://github.com/rocket-pool/rocketpool-research/blob/master/Merkle%20Rewards%20System/rewards-calculation-spec.md), resulting in wrong reward predictions in case of e.g. pending validators.

Additionally, the current rewards documentation (https://docs.rocketpool.net/guides/node/rewards.html#rpl-rewards) is very unspecific about the effective stake calculation.

In combination, this can lead to node operators making wrong decisions about minipool deposits and exits, based on how these affect the effective stake.