FormerlyHemlock
Hero
Simply, because tons of things that sound cool on paper fall apart when executed in real life. When the rubber hits the road you find out how your plan reacts with the variables of a dynamic environment and that always changes things. So doing the solo attack instead of theorizing it gives a legitimate example rather than a theory.
I haven't found this to be true. A 20th level party that looks on paper like it's going to beat the pants off Tiamat using crazy-high Stealth rolls, paladin smites and potions of Fire Giant Strength, will in fact beat the pants off Tiamat using paladin smites and potions of Fire Giant Strength. It's not the "real life execution" that makes things hard and complicated--it's the presence of an unpredictable adversary. But if you're the guy who's planning on playing the adversary, there is zero point in playing things out in a solo D&D game, especially when it comes to dominant strategies like "kite the Tarrasque to death using mounted combat and the Magic Weapon spell."
If you want it to be exciting, you need at least two people: one guy to run the PCs, and one guy to run the Tarrasque; and neither guy should have full knowledge of the other guy's capabilities. (I.e. the Tarrasque should DEFINITELY be house-ruled in some way or other.)
In a game of perfect information, the MM Tarrasque loses. So make it a game of imperfect information.