No way. A team of TB/Rook/Erebus could easily chainstun your whole team for nearly fifteen seconds with only about a one second interruption. Not to mention they could pile on 2500 AoE damage+other abilities and attacks. 3 TBs could continously stun a whole team for 9 seconds and deal 2400 damage, then start over a second later. Teams of four or five? You're removed from the game until they run out of mana or you respawn. Chances are, even if they do run out of mana, you're in no condition to fight.
I highly doubt that any combination of 3 DGs could get an entire team reliably, least of all if they are coordinating to try to prevent it. If any pseudo-guaranteed, infinite combination did happen to be discovered, then I would be inclined to rework the offending skill and keep stun locking, rather than the other way around.
You guys who hate stun immunity must not have played WoW at all.
I honestly don't know what goes on in WoW, but in a real pvp game (Guild Wars) "stun" locking is an integral part of high level gameplay. Chaining knockdowns on monks so that there's no time to cast in between is a very potent method of generating pressure and scoring kills. Every good team does this. Because of that, every good team learns to survive despite it or prevent it from happening altogether. Because of that, every good team learns to shut down their opponent's defensive system so they can go about killing anyways. And so on and so forth.
The point is that as long as its properly balanced, stun locking enhances high level gameplay. And thus should be in the game.
Just because both DotA and WoW have stun locking doesn't mean every game should. It was obviously decided that it would be detrimental to the enjoyment of the game at some point and was countered by adding the stun immunity mechanic. I'm glad they did this, and I'm just pointing out that Foul Grasp doesn't behave any different to the other stuns from the perspective of the victim so I don't believe it deserves to be exempt from stun immunity.
Yes, the attacker is held still for its duration, but its entirely at a time of your choosing, if you don't want to be held still, don't cast it. This reduces the number of situations where it's powerful, but it doesn't reduce its power. Mass Charm has a 3 second cast, during which you're static and vulnerable in the same way, but it doesn't make the skill less powerful than something like boulder roll that suffers neither of these problems.
As for it being only a single target stun, any decent team will be focus firing, so you only really need your target to be stunned, the rest is just icing on the cake. The issue is that if, say, two people on one team are pushing a lane, and they are forced to retreat, using stun locking you can hold one of them in place for so long that they are completely doomed. If their ally comes back to help theres little they can do apart from watch the carnage as the UB shreds it up in melee.
What does this mean more broadly? Pushing a lane against organised stun locking teams is too risky, and makes for a defensive and boring game of grinding creeps and avoiding each other, or more likely a bunch of rage quits from people who don't enjoy spending their time respawning and being stunlocked. Stunlocking horribly gimps your ability to retreat, and that's why it's a bad mechanic, and that's why I think Foul Grasp should be treated like all other stuns.
Paragraph 1: Yes a decision was made, but imo it was the wrong decision. See my reply to Obscenitor, above, for specifics.
Paragraph 3: A decent team can also defend. Teammates can rescue their ally in a wide variety of ways: heal, shield, bramble, silence, interrupts, stuns of their own. If outnumbered and facing DGs that obviously can stun you, retreat earlier.
Paragraph 4: So what you're saying is... you have to be careful when wandering alone through enemy territory? Darn. As for ragequits, I think stun locks are more the property of premade games, wherein ragequits generally do not occur. Unless you are talking about a premade vs pug game, wherein the pug was probably doomed anyways.