Actually, from reading Governor, it would seem that a vanilla "gain a gold" action card should be $4..
Governor weakens the "gain a gold" instruction by having opponents gain silver. That suggests that a vanilla "gain a gold" should cost more than Governor, not less. The fact that Governor offers choices, though, muddies the waters quite a bit, and maybe the best recourse is not to use it as a benchmark at all.
The reason I priced "gain a gold" at $4 is that the other two Governor benefits-to-you mimic Smithy and Remodel, which are both $4 vanilla actions. The fact that the Governor also gives +1 action and give the enemy a benefit and lets you choose between 3 actions are roughly equal modifications to all of these vanilla effects.
I like your mathematical reasoning there, but I think there's a limit to how well that kind of calculation works in practice. And I also think it's flawed to assume that the three choices are supposed to be equal benefits. Consider the following card:
Choose one (you get the one in parentheses): Each player gets three Provinces (two Colonies); or +$1 (+$2).We can quibble about the details if you want, because this is a strictly top-of-my-head example (and I don't know how +$ works when it's not your turn), but my intent was to propose a card where the
margin of benefit between the choices is roughly equal, but the
magnitude is not. We can't assume that, just because these choices are present and both are potentially good choices at any given point in the game, that each component of each choice is equal in power to its counterparts. That is, "gain three Provinces" probably isn't equally powerful as +$1, and nor is "gain two Colonies" equally powerful as +$2. Maybe a clearer illustration of this principle is this:
Choose one (you get the one in parentheses): Each player gets +1 (+2) Cards; or +2 (+3) Cards.And I think we have to chalk the actual Governor card as illustrating this principle. As others have pointed out, a "gain a Gold" card as an opener does something more powerful than any other existing official card does.
There's also something maddeningly non-mathematical about this:
The fact that the Governor also gives +1 action and give the enemy a benefit and lets you choose between 3 actions are roughly equal modifications to all of these vanilla effects.
The value of adding +1 Action onto any given behavior is not constant. As an illustration of this principle, let's take Moat and Duchess, two $2 cards, and add +1 Action onto them.
Duchess' Spying powers cancel each other out, so we basically have a card that offers "+1 Action, +$2." This is almost exactly a Silver, with a few different nuances in the fact that it's an Action card instead of a Treasure card. Certainly it's dramatically inferior to Festival, which is the same except that it gives a second Action and +Buy. So you'd think maybe $3 is the correct cost for a non-terminal Duchess.
A non-terminal Moat, however, is strictly superior to a Laboratory! It offers +2 Cards, +1 Action, AND immunity from attacks. A cost of $6 would be an absolute minimum, and I'd probably make the case that it should really be $7.
What this means for Governor is that even if you establish to yourself that all its options in terminal form are of roughly equal power, it doesn't necessarily follow that they would then be equal in non-terminal form.