I've come to realize that's a somewhat shallow level of game design. I like it when the value is conditional based on the game state. Some stuff is in demand more than at other times. Cool. But there is still a deeper mechanical issue to design around.
The real question is how long does it take to achieve the necessary condition to score points?
The Time-to-Point Ratio
This is a handy metric to reverse-engineer existing games and figure out the balance of your game. For example in Lords of Waterdeep, in a single turn you could collect 4 coins, or 2 orange cubes, or 1 white cube, or 1 purple cube. These resources can be redeemed later to complete a quest which earns you points and other benefits.
So at the very basic level, Domesticate Owlbears could be described as costing 3 turns for 8 points. In addition, you get 1 orange cube and 2 coins, which are individually the equivalent of half a turn, but might be just what you need to complete a second quest, thereby saving you one whole turn. (EDIT: I know technically there is another valuation at play, since cubes themselves are worth 1 point each, but let's focus on one thing at a time.)
In a 2-player game, players get 4-to-8 turns per round for eight rounds. It's tricky to estimate because of Waterdeep Harbor and extra Agents that are acquired later in the game. In general that gives you a baseline of 32-to-56 turns in an average game. Let's cut it down the middle and say you have 44 turns in an average two-player game.
From there, you can reverse-engineer the game to figure out the most optimized path to achieve your goals. Anything that takes extra turns is bad. Anything that saves you turns is good. For example, there is a Building that allows you to spend two coins to gain 2 purple, 2 white, or 1 white and 1 purple. In other words, spend 1-to-1.5 turns to get the rewards of 2 turns. (Depending on if you spend money you had at the start of the game.)
This is further complicated by opportunity-costs, especially in worker placement games, but that's still the essence of engine-building games. This ratio is a skeleton around which to design your own games. How many turns are you spending? How many points are you getting for those turns?
In my next post, I'll dive deeper into the Time-to-Point ratio of the new game Splendor.