The way planetary focus works is very odd, and can lead to very strange build strategies. Take, for instance, a situation where your overall spending is 100% social. A factory planet, focused on research, will generate far for research than a research planet, focused on research. Somehow the factory workers, without labs or training, are way better researchers than those people with all their fancy computers and giant installations built specifically for research.
A change of the economic sliders to 100% research reverses the situation. Suddenly those guys in the factories don't have a clue how to build anything, but ask those eggheads in the labs to build a starship and they'll do it in no time.
People can abuse this mechanic by basically neglecting to build research infrastructure and relying on their machinists to unlock the secrets of black-hole physics. The really odd thing is you don't lose all that much by using focus in this way.
Here's an alternative proposal: when a planet focuses in an area, the other two areas are treated as if they had half their current slider values. The area of focus gets an increase in it's slider value of 2/3 of the 'liberated' slider points.
So, say your slider settings are 20% military, 40% social, 30% research.
Focusing on military for a planet would make the effective settings: 43% military, 20% social, 15% research. Focusing on research would make it: 10% military, 20% social, 50% research.
The 2/3 factor is there to penalize using individual focus instead of using the overall sliders. You could make it 3/4 or even 4/5 depending on what seems an appropriate penalty.