Hauling itself is not a really tedious problem for
players. There's a big difference in game-costs and player-costs. It just takes time and the mere effort of designating the place you want the junk to go.
Instead, it is a problem when you can designate a whole mountain, step away from the game to go make yourself a sandwich, and when you come back, a single dwarf has completed a mountain-top removal project.
What we need is for mining to have some in-game cost that makes it not so easy for players to vaporize whole mountains in an afternoon.
And to go back to your own quote:
evil clouds: deadly and sometimes frustrating, but creates fun.
less stone next update: no more buying caravans with eight million stone crafts, may actually have to work for something.
sieges: dangerous and badly timed, deadly. but add a purpose for the complex combat mechanics, keep in mind sieges are unfinished atm.
I don't think you make a good case, here... Evil clouds have one real purpose: The surface is deadly - dig down, dig down immediately, and never come back up. All caravans and migrants are done for.
Less stone next update affecting caravans? Why? Who bothers making stone trade junk, anyway, when Large Serrated Steel Discs and Dwarven Syrup Roasts can single-handedly buy whole caravans? (And right now, caravans are more effort to scroll through their lists of junk than their goods are actually worth, anyway.)
Sieges, meanwhile, do have a definite fun element: They force players to build their fortress with an eye towards defense. But you aren't showing how this element is in any way really different from mining, as handling sieges typically
is about altering the lay of the land and mining.
Water has a fun element: You cannot directly manage where water is going, you have to work with water in dry environments, and use floodgates or pumps to control when water will be in a given chamber. The indirectness of control makes it more fun.
And yes, there are definitely other systems I want to see added to the game as it becomes more complex that would also be fun, such as sewage systems or ventilation systems which will be fun for the logistical management that we already need to have in managing siege defense and water or magma management, but to say that we should stop at just these things because, arbitrarily, we consider designing the minecart/logistical systems by which we remove rubble not to be fun while designing the systems by which we move water or manage sieger progression
are fun.
The reason why we need rubble is, counter to your expectations, it should
reduce the number of junk items being generated because it slows down the pace of excavation considerably, as each time you mine a set amount of stone, you have to deal with the excess material generated. By the time that you are mining again, that rubble should be dealt with and out of memory in some manner or another.
This isn't the only thing, we should eventually have better cave-in mechanics and requirements for supports or reinforcing walls.
The ultimate goal should be that mining should no longer just be
assumed, that every mineral in a given embark is your Armok-given right to haul out and do with as you see fit - you have to work for your stone, and make do with the resources you can manage to get, rather than simply sitting on tons of waste stone you can ignore forever and probably will never use.
We, as players, should have to look at the geology that we can see, and start sitting down and trying to guess, as real miners do, where the ideal places to mine will be, and where the physical limitations of their ability to mine and set up the support and safety infrastructure will make mining less of a priority or too dangerous to carry on with.