Suggestion results don't always make sense (suggests things that don't increase the prediction)

[hhyyrylainen]

This seems like a pretty major problem (and probably the source of bad nucleus suggestions as well) that when calculating the suggestion numbers the energy prediction is a lot higher than in auto-evo prediction:

A save where this situation can be seen:

Primum_experimentium.zip

[hhyyrylainen]

I think the ultimate cause might be that the suggestion tries to emulate an energy gain calculation but it does it by proxy through population numbers, so a size increase that doesn't result in massive population drop might give a false sense of being better due to higher individual cost increasing individualCost * population while it actually decreases population.