Actually the core of the MP inbalance (the inbalance in such a way that strict rules on expansion have to be determined), I believe is caused by the power void in CK. What I mean by that is the following: There are maybe 20 kingdoms, but there are also very many dukes that can be played and will be played in MP. If we say there are at least 100 feasable stronger entities in CK that ought to have a strong human leadership, let's compare that to EU2.
In EU2 there are many less comparative superpowers than in CK. Superpower should be defined as a nation that alone can dominate the world without too much effort and luck other than regular play. Think of the "minimum player number" for a good long term campaign; we usually want Spain, Austria, England, France, Ottoman Empire (alternatively China and others), and perhaps a few powers such as Russia, Sweden/Denmark, the Netherlands or Poland. Even further we can extend that to Persia, Prussia and the Mugal Empire. The main point here is that minors can be safely ignore because unless played very skillfully and are lucky they will not influence the flow of the game. Actually a game with only 6 players in EU2 can develop perfectly fine; sure countries will be much richer and centralized in the end of the game than their historical equivalents, but there will certainly not be any superflous expansion capabilities.
The greater number of CK entities to correctly simulate the flow of history, or to at least make it feasable makes the AI far more important. Although I'd admittedly like to live to see the day when we will see a 100 player Paradox strategy MP game, the currents setup forces a big weight onto the AI's shoulders. Currently the AI is surprisingly unfit for that role with kingdoms collapsing under its guidance etc. Of course, this is partly derived from the fact that power entities in CK are inherently more unstable than in EU2. In EU2 the AI as of patch 1.08 deals with revolts rather effectively, and there are not very many radical actions that can cause such huge internal threats as in CK. EU2 has stability, but even stability at -3 is very easy to manage unless your country has already expanded beyond its limits and then is already ahistorically successful (once again the AI does not tend to manage that very often).
The feudal tier in CK also lends itself to more problems for the AI than the simple country system in EU2; that is in EU2 one simply has to take a province and its his. Since the CK tier system also leads to the increase of playable entities there are simply a lot more wills to be managed and to be compared against in CK, and it effectively makes for a theoretically more tough game. Of course it's not the human player who ultimately takes the penalty for this; it is the AI which does that. It rhymes bad with the fact the AI is also more important which was previously established in this text.
For any multiplayer game to be successful as such (and not only be a successful singleplayer game in a multiplayer environment), the interaction between the humans are the most important part of it. This is obviously the reason to play MP since there is always the option to choose SP over MP. I believe this concept currently is lost a bit in CK with the reliance on the bad AI, and I do not have any direct suggestions about how to improve this balance.
That's why I have the rules here in the first place, in order to encourage player interaction. I have a pretty firm belief that the game otherwise would devolve into pure AI bashing, although it so far has been quite similar to this phenomena as we expand inside the allowable boundaries. I am quite prepared to change these rules if they don't work in accordance with the game mechanics, but I prefer to keep as many in a workable order as possible with the above reasoning.