Zagfel, two games. A TBT and RTT of your choice. I don't care which. Take two melee units with the same movement and run one past the other while it's under attack. What results do you get?
Okay, well... if your going to let me choose the terms... Assuming the melee units start within contact of one another (because otherwise, assuming the speeds are even, obviously one is not going to catch the other), I'll take a goblin sapper chasing a peon in Warcraft II and an axman chasing a tank in Civ II (both melee in the sense that they have a range of one hex).
Obviously, I'm joking since those aren't exactly "fair" examples, but I'll address the crux of your criticism below.
... In the turn based games, the guy that tries to run past his enemy dies, fast. In the real time games, he gets to where ever he's going with a single hit being placed, at best. There isn't any brainpower in this exercise, no logistical feat has been accomplished, it's pure idiocy. A catastrophic mistake in any remotely realistic situation is instead a highly rewarding method of play. You don't need to draw the enemy away from your objective or defeat them, simply ignore them and attack it first. As long as you have the necessary firepower for the task, you're all set. It sums up most of what passes for tactics in the typical RTS game, it's depressing.
I'll break this down into two objections.
1) It isn't realistic.
Certainly. There are numerous historical examples in which commanders have decided to pursue some objective rather than directly engaging the enemy. But more importantly, even if it weren't, so what?
2) It doesn't require thought.
It does. The most likely scenario is that the attacking unit is attempting to perform some sort of early-game (probably economic) harassment. In order to even attempt to accomplish this in the first place, the attacking player has to make a series of potentially game-breaking strategic judgments: 1) determine whether the loss of the unit is worth the benefits of pressuring her opponent, 2) determine whether the benefits of pressuring her opponent are worth the costs to her own defense from counterattack, 3) determine whether the benefits of building the infrastructure for producing that unit are worth the benefits of building economic infrastructure instead. In addition, there are a number of tactical skills that need to be exercised, including: 1) scout the opponent's base to determine the location of the objective in relation to her defenses, and 2) move the unit to a location where it can rush past the defending unit without that opponent noticing and countering.
And raise you another.
These are exactly the same sorts of decisions that go into early-game harassment in TBS games. The only difference is the graphical representation. ie. Player A's unit decides to rush into B's base and attack a target, while accepting that eventually B's units will kill A's unit, versus Player A's unit deciding to attack B's base to weaken some target while accepting that either B's defending units will kill A's unit either immediately or next turn.
I am sympathetic to people who hate the frenetic pace of most modern RTS's. But that doesn't mean that I don't recognize that they involve significant cognitive skills.