On the other hand, luck can sometimes render even the best tactical play meaningless in a "swingy" system.
To use your 3e giant example, it's still possible that even if you tumble back to avoid a full attack, the giant will charge, crit for double/triple damage, and kill you anyway. Chance can prevail over the use of sound tactics.
Note that I'm not saying I think a "swingy" system is inferior or any such nonsense. IMO, "swingy" isn't objectively better or worse than "grindy"; it's just different.
In 4e, the difference between good, medium and bad tactics is about 1 resource unit (1 daily, AP or a few HSs over the party). Yes, adequately bad tactics will cost more, but that takes work. Yes, it is possible to adequately balance combats so that good tactics are vital, but it is hard. Good tactics, in general, just aren't that important. You have a brief moment with a tactical advantage, and then you are returned to your regularly scheduled grind.
In 3e, the difference between good and bad tactics is can easily be one of going from "if we are very unlucky, someone will die, it would take epic bad luck for a TPK" to "unless we are very lucky, someone will die, and we have good odds of a TPK". Yes, if the giant charged and critted, someone could die with good tactcs, but the party still had reasonable odds of victory. On the same roll however, with bad tactics he would have Critted, Cleaved, and then continued his iterative attacks into a secondary target, setting up a rapid TPK. In this scenario, good tactics made a loss-less victory probable, and even in the event of bad luck, prevented a brutal TPK.