Its worth noting too that simply increasing the bonuses from buildings is not going to help. If a is the number of specialised tiles available, b is the bonus from specialisation improvement (ie factory/lab) and c some constant representing your food, and f is some number representing other bonuses such as starbases, tech bonuses, etc. then a simplified equation for a specialised planet takes the basic form of:
While, a planet split 50/50 and running research project has a simplified equation of the form:
(c+(1+ab/2+f)/2)*(1+ab/2+f) = c*(1+ab/2+f)+((1+ab/2+f)^2)/2
It doesnt take large values of a,b and f for the second equation to completely dominate the first.
It is that latter term that causes the balance issues - you essentially get the square of your starbase bonuses, tile specialisation bonuses, tech bonuses, etc. Assuming you upgrade both factory and lab tech paths - which you need to do anyways. Note I've only included in "f" bonuses that are applied to a planet in an inherantly diversified way. ie, a given starbase provides a bonus to production and tech. Each lab bonus in the tech tree usually has an equivalent. etc.
The problem gets even worse as you apply more bonuses - precurser relics, ideology bonuses, planet bonuses, race bonuses, etc etc.
In lategame (nearly turn 300) im seeing my unoptimised diversified planets outperform my specialised planets by at LEAST 2:1 - and I spent the whole game focusing on optimising specialised planets. Im fairly sure if I went back and optimised my diversified planets they would outperform by at least 3:1