Whilst my comment was semi-trolling, the fact is that the BG games were dumbed down by the move to RtwP; the original D&D gold box games were glorious in comparison, and possibly because Pools of Radiance was one of my first proper CRPGs, I still regard that as being the proper way to translate D&D to the PC.
Whatever people try and claim about the rtwp set up, its simply not as good as turn based for ultra-tactical play
I'm replaying Pool of Radiance currently, and while it held up far better than I thought it would, tactics are pretty sparse.
Most battles are won through a rudimentary formation of "tough dudes up front, squishy guys in the back" and there's little room for tactical flexibility of any sort. Sure, you throw a few Sleep, Stinking Cloud, and Hold Person spells here and there but use of these spells pretty much fits neatly into the aforementioned tactical scheme. And most of the tactics involved with terrain are pretty rudimentary; most chokepoints are basically gifted to the player, like the one in the enormous Sokal Keep brawl which happens to be right next to the player party spawn point. More "advanced" tactics border or engine/AI exploits. Things like using weaker, lower THAC0/damage enemies as "walls" to trap stronger combatants behind, camping on troll's spot of death to prevent regeneration, or using the hardcoded reticence of enemies to step through Stinking Clouds to your advantage all seem on the level of door abuse in the IE games. The big difference is, IE games, with more abilities and thus greater tactical flexibility, don't hinge on exploits and they aren't a large proportion of the overall tactical pool.
Variance, while present in all D&D games, is far more prevalent in PoR. The biggest offender is the initiative mechanics, something that plagues every turn-based D&D cRPG. Rolls with a largely random component are a terrible way to determine sequence of actions, especially in a system in which the first mover gains a tremendous advantage. Battles can easily be won or lost in a round, with one (un)lucky roll...it seems silly to have such an advantage hinge on random rolls. But that's how it is in Gold Box games, ToEE, or KotC. Sometimes your rolls are bad, and one or more party members are out of the fight before you even have a chance to act. That's anti-tactical (to say nothing of the fact that order of actions is inaccessible information in the Gold Box games).
And the there were the little things, like healing/turning lacking the deterministic outcomes of the IE games. In Baldur's Gate, one knew how much curative magic would heal, clerics would always be able to turn a certain quantity of hit dice of level-appropriate undead, allowing the player to make informed decisions. In Gold Box games, especially Pool of Radiance, not so much. Turning undead is extremely random. Once a 1st level cleric turned about a dozen skeletons and zombies whereas a 3rd level cleric failed to do anything in four consecutive encounters. Healing is pretty bad as well, with Cure Light Wounds healing between 1-8 points of health and seemingly, though there may be some bias here, tending towards the low side (lots of ones). This means healing in combat is dicey, at best, because one can't count on a significant quantity of health restored to justify the time/resource commitment.
Also worth mentioning are encounters in which enemies have extremely powerful abilities of which the player party has no counter besides appeals to the dice (be it in saves, to-hit rolls, or fortuitous attack rolls). Poisonous foes, level-drainers, and such have no playable counters available to the player, meaning that fights with these baddies are often out of the player's hands. Good strategies against these foes don't ensure clean victories, which spoils any real tactical feel.
My recollection is that some Gold Box games were a bit more tactically interesting, but I also recall some of them descending into total silliness (Pools of Darkness a.k.a. Delayed Blast Firebal: The Game). And as I'm still playing through PoR, and am likely to play some more SSI crawls, I'd be interested to hear of any cool tactics I may be overlooking from any Gold Box mastermen.
(e.g. wizard duels in 2nd edition, which were all about countering the other guy until one of you got unlucky; I recall a couple of tough fights in BGII that were basically set up around the wizard duel contests [something to do with a lich rings a bell] but you could really see the strain being imposed mechanically by the rtwp set up).
There is nothing "lucky" about dealing with spell/combat protections, as there's no randomness with any of them, save relying on Dispel/Remove Magic (which are extremely powerful and strip away most all magic effects for the low cost of a 3rd-level spell slot). Once the player has a strong grasp of the underlying mechanics, selecting a correct anti-magic scheme (in most situations there are multiple good lines of play) is simple, and more clever players will also realize that many defensive combinations have holes in them that can be exploited.
The biggest complaints I have against the sub-systems of magical protections are that the means of participation were not well distributed among the classes and that it interacted poorly, at times, with certain creature's innate immunities (e.g. Rakshasa being immune to spell levels 1-8, Liches being immune to spell levels 1-5, things immune to normal weapons casting Protection From Magical Weapons), but it was remarkably free of silly variance.