Yeah, I'm not one to use realism as the reason to justify a feature. I do use it as a reason to nix a feature, though.
There's Nothing Wrong with "Preventable"
There are lots of things in Civ that are completely preventable. You can prevent an opponent from taking a city, or breaking a critical road, or prevent your empire from going bankrupt, or prevent your city from starving, or from falling into civil disorder. Some of these are easier to prevent than others.
Does that fact mean these preventable problems should be removed from the game? Of course not. They exist not to penalize the player, but as negative reinforcement, to encourage the player to play differently. Don't blow all your money. Build larger networks of roads. Have a strong military. Build temples. Pay for luxuries.
Of course, it takes at least an intermediate player to do all of these things at the same time, even if you add "keep your empire unified". And slightly better than intermediate players might start to take risks. "I'll blow all my money, I'll just try to get by with 3 gold for now." Or "I'll just have the one road for now. Right now, I need to irrigate." Or "Sure my people are getting unhappy, but I need to invest in science."
And once an opponent comes along, they can suddenly be kicking themselves. "I should have reinforced that road!" "I should have kept some emergency gold to rush production!"
I would imagine secession to be the same way. Easy to prevent. Something you might take a few risks with. And something where if you took too many risks, you might find a set of bad circumstances forcing you into a tough situation.
Not to mention that I think it should be easy to recover. If you can mow them down with your army relatively quickly, the rest of the cities should fall to you automatically. (That is if your military isn't already spread too thin, and there's no opponent who hates you enough to offer a protectorate to the new nation.)
No Micromanagement
I don't imagine it as something that requires a lot of micromanagement. The horror story is it's a lot like civil disorder combined with culture flipping. You're checking every one of your cities every turn to ensure that your happiness scales don't break a threshhold, and your culture isn't too impoverished.
In reality, I imagine secession being a replacement for civil disorder and culture flipping. One unhappy city won't flip. One culturally impoverished city won't flip. One city with both won't flip -- it may just have lowered productivity. Secession would only happen if there is an entire region (3 cities or more?) that is both culturally improverished and unhappy. The cities would have to collectively pass a threshhold to seccede.
Moreover, I think this should be obvious from the overhead map. Instead of a little "civil disorder" fist flicking on when it's already happened, you should see ratings on your cities, from Blue meaning "We love you" to Red meaning "We hate you". These would happen gradually. You might be able to live with Green -- "we like you", or even yellow "we're getting by".
There's no micromanagement involved. You're not doing anything on a turn to turn basis. You address the crisis when you feel like it's starting to get out of hand -- and with the regular channels of cultural improvements or happiness improvements. And it's easy to see how close you are to a crisis.
Big Picture
I do think that this ties into a lot of other things. I think you'd need to improve culture so there are more intermediate payoffs to pursuing it than just "cultural victory" and "city flips". That way culture seems less like a nuisance and more like a real strategy that you'd want to pursue. (Consider Cultural great leaders, culture enabled improvements and wonders, and so on...)
I also think you'd need to improve espionage to not depend on large lump sums but small amounts of constant investment. That way it becomes a real strategy you pursue, rather than a split second decision. Not to mention that I think other factors like culture and your respect around the world should have an impact in your ability to recruit spies. (e.g.: If foreigners like you, it's easier to attract spies.)
But hey, I'm talking a lot to convince somebody who is highly skeptical. There are those who have faith. Not necessarily because they know how it will work, but they believe that it CAN work, if implemented well.
The key to a complex strategy game is a hell of a lot of balancing.