There is a problem with Lao-Tzu's definition of the northern border - since the origin of the river Sava is way north in the Alps, near the Italian-Austrian-Slovenian border, and since Sava flows from northwest to the southeast there, you'd only have a triangle-shaped part of Slovenia as part of the Balkans, which is not quite common.
If you, however, place the point in upper Sava flow where you start drawing the line down to Rijeka further down the river, you quickly realize that there is no longer a part of Slovenia on the Balkans.
If we dismiss the idea that Slovenia is on the Balkans, then the issue of Croatia arises. The continental, panonian part of the country is on the north of Sava, and Istra is to the west of Rijeka, so only part of Croatia that would be on the Balkans would be Adriatic coast. An informed sociologist or historian would easily refute the idea that Dalmatia belongs to the Balkans (instead, it is a typical Mediterranean region). Considering how pure geographical criteria is not ultimate (Greenland being considered part of Europe, or Turkey, or Israel), on these grounds 90% of Croatia wouldn't be on the Balkans.
But how can Bosnia be on the Balkans, when Croatia which surrounds it isn't? :-)
And the inclusion of the whole Romania also begs the question, why isn't Hungary included, when there are parts of Romania that are farther to the north than Budapest?
The southern border is also moot, since most of Greece is one of the prime examples of a Mediterranean country. Greece is also in the EU, which must make it vastly different than other "typically" Balkan countries. (Or not?)
Therefore I am of the opinion that the borders of the Balkan peninsula are impossible (or at least very hard) to reliably determine geographically, and that the issue is a volatile, political one. Then again, maybe I'm just showing bias, take this writeup cum grano salis :-)