The current formulas for some game mechanics are far too simple and one-dimensional. This is something I would really like to see changed in the future. The biggest culprits are probably research, economy, population growth. There are probably others equally guilty but I'm not familiar with all the formulas yet to judge them one way or another.
population
This had to be brought up elsewhere. This linear +.1 pop growth + % modifiers per turn is absolute FUBAR.
Population growth should be weighed with the current population and the population capacity. Growth is maximized when the population is at half the capacity, and it should drop as it gets lower or higher then this. Most turn-based space games follow this, and it adds depth and strategy to colonization. +growth planets become 'feeder' planets, where populations are scooped up and transferred to frontier planets.Planets with minimal populations will experience minimal growth, their development will depend on receiving more population from elsewhere or they will take forever to grow.
To minimize micromanagement, allow players to transfer populations directly from the planetary screen. Doing so automatically spawns a transport which moves at the rate of the best single engine available. Transports can be intercepted by pirates, and can not be directly controlled (they are private craft).
Research and economy
These are global mechanics, but all the bonuses are given on a per-planet basis. This makes all global modifiers to these stats extremely weak. Racial bonuses and Relics should be giving a FINAL modifier to research and economy which multiplies the SUM of all planets output. We have far too many additive multipliers in this game, and I'm only picking on research and economy because they are the most obvious.
When my race has a 10% research bonus, it should actually equal 10% more research. The current system adds a 10% bonus to existing planetary bonuses, which effectively gives us a < 3% bonus even early-mid game. Economy bonuses are the same. They are completely insignificant. By late game a 10-20% bonus can effectively provide under 1% research.