I have used this tactic, usually with my Knight, with varying degrees of success, I can't really speak to far past this, as someone who fluctuates between 1200-1300. But getting the king out into the open and stopping your opponent from castling can be very beneficial especially if you have some decent development. I would be more interested in what someone with more knowledge and a higher rating thought on the matter though. Hope this helps.
Sacrificing a Bishop to draw out the King

That's why I suggest using your bishop- the knight tends to be more important when playing through after this strategy is initiated. It would be great if some more experienced players would comment on this..... I've used it on a 1694 and 1926 ranked player- I won against the 1694 and I'm still playing the 1926 (we're about even but I'm like one point behind).
Thanks for the comment!!
It is a pretty bad move imo. Not to say I haven't done it before... but the very few times I did it was because I had been playing sloppy and got backed into a corner. Once was a game I was half paying attention do, and opened with my Knight to set up an Indian defense without even looking to see that my friend had opened with his King pawn, which he then pushed forward to threaten the knight... I moved the knight forward and he kept chasing it until I got tired and sacrificed him on the KBP square. I felt kind of stupid for even having to be in that position in the first place (on the plus side I won the game with a quite nice fork on the queen).
Anyways, moral of the story is, if he is going to die anyways, you might as well let him die for a reason, but it should never be your plan from the get-go.

It doesn't usually work above 1400 level unless it's a really well-thought out sacrifice. Surprisingly many people will just sacrifice a knight for the f-pawn and expect to see a winning attack, when they have simply wasted three moves just jumping the knight to its death.

Okay, let's limit this to the bishop sacrifice only. Like I said before, the knight is more important in this situation. Also, I'm suggesting this in the very early setup. That is, taking the pawn when it hasn't moved yet, and only then when the king would have no choice but to retaliate to the diagonal. This is, after all, a relatively limited tactic.

Eh....... I guess. I'm not familiar with all the etymology of chess moves, etc. It just seems to have worked well in a few of my recent games.

Screenname "elithepirate," the Former Mississippi State Open Champion, Former USCF Expert, Twice Top Expert in Tennessee Open says:
The bishop sack againest a castled king, normally on f2,h2 or f7,h7, works if and only if supported by other pieces. A solo strike is impressive but the defending king should be able to escape to safety. A knight or queen form the most natural supporting pieces.
How's that for some expert data, eh?

I prefer the bishop sac on h7 after the king has castled short or allowing the bishop to be captured on g5 by a pawn on h6 when a pawn on h4 can recapture.

Honestly, it's often much better to attack a castled position with this kind of sacrifice in the opening/early middlegame. I'm talking about a greek gift sac - bishop takes h-pawn after the black king castles kingside. It requires some backup (knight f3, queen's bishop and queen have diagonals to support), so once again it can't be done automatically.

using the bishop sacrifice in the way you described seems to work only if it causes panic like it might in a 3 minute blitz game. It can lead to a strong attack but with a cool head and a few minutes i have almost always found my way back to an equal position after my opponent has played the sac, leaving them down a minor piece for the pawn. It only takes a few tempi to castle artificially or find a safe spot for the king to sit.

There is a chess saying - the threat is stronger than the execution. Sometimes it is good to place pieces at f7, for instance. A perfect example is something like the Muzio Gambit, which is as follows...
Now, the idea behind the threat being stronger than the execution. For instance, let's say you have a bishop aimed at f7. But you don't just sac it because you can. What if Black has plenty of defensive options? Then you will be down a piece, and as Petrosian said, the king is a fighting piece! Sometimes Black can use his king in the battle to protect itself! If you look into some of these "Greek gift" games, you will see that Black wins some games with an advanced king. You can continue to sac the bishop if you want, and using a computer might help you find cool tactics, but to plan to do it every game? There are so many defenses. For instance, what about in the French Defense? You can't hit f7 with the bishop on c4, so as they mentioned, you have to "hope" for the opportunity to play the "Greek gift", and you really aren't promised that Black will castle kingside.Here is your French Defense diagram...
So you may want to play openings that allow you to play for aggressive attacks, but understand that Black can play more solid defenses, like the French and the Caro-Kann, where Black plays ...c6 on the first move instead of ...e6.
This may be an old tactic (I'm not really sure- ), but using a bishop/ knight (I prefer bishop 'cause it actually forces check) to take the pawn to the right side of the king and let the king take your bishop seems to be a good tactic. It may not be an equal exchange, but it cancels the opponents castling and if played well can draw the king even further out. Thoughts?