There are two kinds of political borders in the world: "Switzerland" borders and "Colorado" borders. The former kind is irregular and complex, and results from centuries of wars and political maneuvers which lead to innumerable, incremental border adjustments. The latter kind is simple to draw and remember, and occurs when new political powers come and wipe out the previous political order and need to make new borders from scratch.
The borders between Israel and its neighbors are generally in the second category, as they result from the imperial struggles of England, France, and the Ottoman Empire in the early 1900s. Thus the Israel-Sinai border is a straight line (with a few kinks which actually result from errors made drawing that line), and the Israel-Jordan and pre-1967 Israel-Syria borders essentially follow the Jordan River and Arava.
The Israel-Lebanon border is different. It is surprisingly complex and seemingly arbitrary, given that it was the result of a simple partition treaty between Britain and France. What possessed the map-drawers to add so many bends and twists?
I think the answer can be found by looking carefully at a topographic map, and remembering several considerations which must have guided the mapmakers. In particular, there seem to be two critical locations at opposite ends of the border which determined the rest of the border: Rosh Hanikra and Dan.
Dan, and the whole "finger of Galilee" that it's at the end of, are important as a water source. It's known that the British wanted the Kineret's water sources, as much as possible, to be on their side of the border because without this water it would be hard for the state to be viable. The springs at Dan, and the rivers which descend towards Dan from the Hermon, form a large fraction of the Kineret's water sources. So the British made sure the border went just north of Dan.
Rosh Hanikra seems to be arbitrarily chosen as a boundary between the French and British. But it's such a notable geographic landmark that it's easy to see why they picked it, and not some other place nearby.
What about the long and squiggly border in between Dan and Rosh Hanikra? What I recently noticed, and which impelled me to write this post, is that the border very nearly follows the watershed between Israeli and Lebanese. rivers. Basically, for each piece of land, imagine where rain that fell there would eventually flow. If it would flow into the Mediterranean Sea at a point north of Rosh Hanikra, then the land is Lebanese. If it would flow into the Mediterranean south of Rosh Hanikra, or else would flow into the Kineret, the land is Israeli.
Now, if you actually try that, you'll find that the division line is close to the current border, but even squigglier. And so I believe that after basing the border on the watershed, the French and British went out and smoothed it out a bit just to make it a little more reasonable. The largest consequence was north of Metula, where the Iyun river (which flows in to the Jordan) became part of Lebanon. But there are other such adjustments all along the watershed line, in favor of both Israel and Lebanon.
But look at the sharp bend in the border just north of Tzefat, and look at the pattern of wadis in the same area, and the basic correspondence between border and geography will become perfectly clear.