A hub which serves n destinations can theoretically offer a maximum of n*(n-1) connecting city-pair routings (this assumes of course that the timings work and that the detour factor via the hub is sensible for all the city pairs, assumptions which are not the case in reality but which set an upper bound).
If half of those n destinations are reallocated to a second hub, each hub will now offer (n/2)*((n-1)/2) city-pair routings, i.e. a quarter of the previous total. Two hubs, so altogether there are now half as many city-pair connection options as with one hub.
Of course there are many reasons for limiting hub size (all manner of diseconomies of scale, delays, political pressures, etc.) but in terms of pure connectivity, the bigger the better. "Balancing" traffic between hubs means spreading the passengers more thinly, which means using smaller aircraft or lower frequencies (and there may be competitive advantages to this, but it's a different argument to "boo hoo, yours is bigger than mine, that's not fair.")
Consider what this means if you are a small airline which wants to start a new service to one or other hub to feed/defeed connecting passengers. All else being equal, you'll choose the largest hub
because it gives you the most connections. Your new service will make the largest hub even larger, so there'll be a positive feedback effect.
Of course, not all the passengers in the hub airport are connecting, so a big successful hub typically has a very large proportion of origin/destination traffic too. I think that's one of the factors that's always made it difficult for Swissair/Swiss (and to a lesser extent KLM, and others) to build their hubs to competitive levels.