Cracks in general can be stopped (or slowed at least) by drilling a hole at the end of the crack. This stops propagation, provided the metal grain boundary (that cracks follow) has been drilled out. In layman's terms, the drill hole has to have completely removed the 'end' of the crack, and the edges of the drilled hole need to be deburred (smoothed) because any sharp edge/corner is known as a 'stress riser' and therefore a location a crack can propagate from. Other than straight-up abuse, cracks can form from inclusions (impurities) in the metal that compromise the integrity of the alloy.
So, if you see a cracked cymbal that has been re-drilled several times trying to stop a crack, you know that either they did not properly stop the crack or the stress is so high that the crack can use the drilled hole itself as a propagation point- i.e. it is too small a drill hole. 1/16 inch is probably too small to effectively stop a typical crack, and 1/8 is probably more than sufficient in a thin cymbal. Thicker cymbals may need a larger drill hole than a thinner one, due to reduced flexibility of the alloy (because of the increased thickness).
Thicker cymbals are less flexible as mentioned above and therefore the area of cymbal around the crack (or up to it) may have leverage to continue a crack. Imagine a clean piece of cardboard- it has a certain stiffness to it. Now cut it in the middle across the center, but not all the way to each edge (simulating a crack) and it is very easy to fold in this location flapping the cardboard up and down. The flat un-creased cardboard is like a lever on the compromised area of the cardboard. The same thing happens in the cymbal.
Cracks with the groove in the bell area can be very difficult to stop for this reason- the entire rest of the cymbal is acting against this weakened area, that has support of the stiff bell and immovable cymbal stand, so even with (well done) crack stopper drill holes, the crack can continue to propagate through the hole because the leverage here is high when the cymbal is struck. Plus the vibration of the metal itself adds to fatigue which can cause cracks all by itself .
The lathe groove is also a propagation point for a crack, and I have seen cracks with the groove that shoot straight out from the crack end radially, to then take a sharp turn to follow another lathe groove. (in this case it was just played and no attempt was made to stop the crack)
So what I'm saying with all this is it is a case-by-case process and a drilled crack may live for another 100 years and it may not last beyond the next time you hit the cymbal and is dependent on many factors.
Cutting an edge crack completely out can have two benefits: 1) it lightens the cymbal where you cracked it, so on the stand that 'weak' area will be turned away from you. 2) it completely removes the propagation area from the cymbal- but be careful- I see a lot of crack repairs where a 'square' section is removed with sharp corners to the cut. It is FAR better to remove the cut with all rounded corners (even an arc), or you set up a new stress riser for vibration induced cracks to form at the sharp corners of the removed area and these tend to run around the cymbal following the lathe grooves.
Long winded but I have a bit of crack and corrosion knowledge and have an understanding of what is going on here (to some extent anyway).
