Before we continue with this discussion, how about we actually understand how it works first? No insult is intended here, please don't read it like one. The game mechanics implications are not very transperent unless you examine them thoroughly. This might actually be the bigger design flaw then the actual mechanic.
Last I checked, Coercion applied a penalty to raw production that stacked multiplicatively with the various bonuses to raw production from economic starbases, improvements, and high approval.
This is not true. I think what deceived you here, is the double penalty -% morale and -% production. Just start up a new game and look for yourself in the planet view how it calculates. The production penalty is additive with approval and starbase boni. Because you get penalized to both morale and production however it can sometimes seem like it was multiplicative to the other boni.
So what difference does this make? One difference is, that morale actually is a soft counter to coercion. If your morale is high enough to offset the -50% you only get the production penalty. With max approval and 5 eco starbases this means going down 33% of production, not 50% or more. The next thing it does is change the wheel positions where one output (in one category) is maximized. If you start with a civ which has no content bonusses you get your maxed output at about 70%. If you play a civ starting in the torian system with content +2 your maxed output is somewhere around 80%.
I will have to test a little more, maybe for a specialized research world a bureau of labor and high morale bonusses/ lower pop is actually advantageous.
To summarize: This does not necessarily make coercion an unflawed game mechanic. Even if it did, it would still be bad design, when people like joeball - who knows how math works - get it wrong. Heck, I wouldn't have looked twice if it wasn't for this discussion, I had it wrong, too. If your mechanic offers depth (which I think coercion actually might) but the majority of players doesn't recognize it, it is a failed design. But none of this changes my opinion which I stated earlier:
since coercion I micromanage less. For me it made the game better, regardless on whether the mechanic is great or aweful. It was a step in the right direction in my book.
I will continue to paly with the numbers on coercion a little, to see if there actually is some depth to it, or if its numbers are wrong.