I still maintain that the GÇ£pay for number of targetsGÇ¥ logic is wrong-headed GÇö no matter the base cost and any diminishing returns, it only ever means that small targets will be picked on and that dec-shielding will become the standard.

Make it a relative measure: you pay for number imbalance.

abs( ln( attacker size / target size ) / ln( size multiplier ) ) +ù imbalance cost + base cost.

In other words, for every [size multiplier] times larger or smaller the target is than the attacker, the cost increases by a factor of [imbalance cost], with a minimum price tag of [base cost]. This gives you a lot of variables to play with: how cheap will any war be (base cost)? How much do I have to pay to bully a small guy or annoy a large guy at the Jita undock (imbalance cost)? And, most interestingly, what actually counts as having an GÇ£unfair numerical advantageGÇ¥ (size multiplier)?

E.g.
A size multiplier of 1.5, imbalance cost of 50M and base cost of 5M GÇö for every 50% increase in the size difference between target and aggressor, the war becomes 50M ISK more expensive with a minimum cost of 5M for perfectly equal sizes.

-+ A 10-man corp attacking a 1-man corp (or vice versa): 289M ISK.
-+ A 10-man corp attacking a 20-man corp (or vice versa): 90M ISK.
-+ A 3,500-man corp attacking a 5,000-man corp: 49M ISK.
-+ A 5-man corp attacking a 5,000-man corp: 857M ISK.