You don't follow?
OK, I will say it as simply as possible.
We have 6 incomes.
I concur that you have X incomes (6 for this theoretical), the issue isn't about the number/size of your incomes, but rather how much of that potential maximum income you will be able to apply. The one-regent realm, while having less income can focus 100% of their income with one regent's decision. The multi-regent realm needs co-operation to be able to focus 100% of their income. Friction will make achieving the 100% focus only a theoretical possibility for the multi-regent realm (Chapter 7 of Clausewitz's On War is a good read here).
Actually the comparison is of realm A - income split over multiple regents and identical realm B - all income in the hands of one regent, because the problem noted is what happens when (generally board/computer-gamer type players) simply glomph every holding in their land. The single regent has - at least - the same total income since the holdings and province levels are identical in the sample sets.
The inevitable efficiencies of cutting down on the courts of other regents then gives the consolidator a very large surplus to apply as they wish - most BR mechanic systems tend to have a "critical mass" point in income generation and ROE is no exception - a domain that has base expenses of, say, 21 (4 for holdings, 6 for court, 3 for regent actions, 4 for at least 4 court actions) and income of 24 is eking its money out and likely has only 4-5 units; one with similar expenses but income of 36 is not 50% richer in practice as the gross suggests, but has 5x the surplus wealth to spend on military, buffing actions, etc - it's a serious temptation for the computer gamer who struggles with the concept of "dude, that's
people not just resource-generating units, and its r-o-l-###ing-e playing not r-o-l-###-l playing"
They key risk with the tactic is the clean-up problem, short of a major ret-con or similar the game is routinely left with a very very broken set of domains which damages the game, quite possibly to breaking point, the fact that the player tends to have quit / been kicked out is therefore fairly redundant at that point.
For the record I don't believe that a numbers based rule is needed, but rather the statement within the rules that domains that contain unrelated holding groups are more likely to experience internal friction (events, etc). This makes it clear to players that grabbing everything is likely to result in role-play challenges and leaves the reaction in the hands of the DM rather than a mechanic.
Any time - any time at all - that there is a base effect - positive or negative - the mechanics should reflect it, so that the DM is freed up to focus on the ones in need of particular attention, fun, etc. In practice different military units would cost more/less as pay rates vary, minor lordlings gain/lose ambition, etc - in the game the DM doesn't have time to consider how every unit is composed, their situation, etc, so they need a mechanic to set a "base", and the DM then intervenes when it's special or the player contrives a situation and asks for it to be recognised. The only real restraint on bread-and-butter rules should be clarity and simplicity - it shouldn't be fear of setting a base or outlining a norm.