The idea is that reach type of holding has particular interest, goals, etc - so two holdings of similar types work well within the same organisation.
That's a silly idea. Think about it. A temple of a god with a focus on crafts would be expected to have guilds, and one with a focus on trade or wealth would be expected to have trade holdings, etc.
The concept to encapsulate in a rule is "how many potential disputes on what to do or who should do it will occur in the domain", and this gives rise to the question "will there be more or less potential disputes if there are twenty law holdings present in the domain, or 10 law holding, 5 guild and 5 temple" - the different holding types are more likely to have different needs, employee ambitions, etc than similar holdings so in the latter case the answer is generally going to be "more" - it
might be the same, but if a general rule is wanted than it goes "more", it's no different to saying a lawful domain finds it easier to raise stability than a chaotic one.
The "minor side show" issue is one that I'm aware of, I'd expect less problems in a 90% temple + 10% guild domain than in a 50:50 scenario where no member knows if the next head will be a priest or a merchant, or if the year's surplus will be spent on a cathedral or bonuses. I might be persuaded to an "x"% rule with lower divergences ignored, or that if there are less than "y" holding levels you'd ignore the problem, but I'd want to see how the book-keeping worked in practice.
And if you really do think that the supposed difference in goals, etc. would affect things (I do not subscribe to that point of view) then the DM should create those tensions - that's his job. But to have an arbitrary rule that basically says "If you do things out of the ordinary you will be penalized" is bad DMing, and bad rule writing.
If the player does something that will significantly increase tensions between their holdings then yes, that's for the DM to sort out on a bespoke basis (whether between different holding types or similar or even different levels within the same holding), but I wouldn't want to do that routinely, I'd rather have a simple stated rule than an arbitrary "the DM noticed this turn and thwacked me" approach - plus, by stating a rule up front people can plan for it which avoids the appearance of capricious smacking down of the unexpected.
If the ruler tries to grab everything, let him. he's just made a bunch of enemies who have a common cause against him. You should not write rules with specious explanations to prevent people from doing things differently.
The suggested rule doesn't prevent, it just states the cost up-front, people wishing to grab multiple types can still do so, the suggestion is merely a minor increase in DC to stability checks - you can come up with whatever fluff explanation you like if you don't consider that a merchant might have different ambitions, etc to a noble or a priest in how the profits and influence of the holdings should be spent.
Rules are written to handle broad-brush common areas to reduce need for fiat, discourage arbitrary decisions, and encourage strategic planning - but also to encourage certain styles of play. BR was originally designed to give each class a holding type, historical counter examples (like Monks selling wool and wine, nobles investing in business, guilders buying titles, popes being princes and the like) were circumscribed to aid game balance (if taking things too realistically I suspect that theocracies would be endemic given historical church power and the addition of tangible proof via magic). So in BR the divisions between land, church, guild, etc are greater than "would be expected" because that's the way the original designers built built the holding types, since I don't want to see the guilds crushed out of existence, or feel compelled to conquest, some sort of recognition in the rules to encourage "stay within the bounds" type play seems appropriate, although I'm a little concerned about how far the suggestion goes. Where it works less well is "common pairs", i.e. province+law, or law+manor, or guild+trade, etc. Possibly something to think more on - the issue may in reality only be on crossing guilds since RP protects temples more and sources are less "profitable" for rulers, but in Cariele for example the risk is the reverse of the norm.
