You can measure the voltage, with a DC voltmeter, between ANY two points in a circuit and there can be a difference. Just because there is a difference does not mean there is a voltage drop between those two points. I can measure a voltage difference across a capacitor when there is no DC current flowing

You specifically said the points were the collector and the emitter, not two random points. I do not believe most people would have a problem with saying "the voltage drop from collector to emitter of an NPN transistor in cutoff is X volts." If you wish to restrict the use of "drop" only to voltages measured across a resistor, that's fine, but many other people do not adhere to that usage. However, let's leave this as a semantic disagreement, not worthy of further discussion.

 

Clearly there is a whole lot of confusion here about a transistor that is cut off. There will certainly be a voltage because it is a lot like an open switch. While the current is not flowing there can still be a voltage. Not because of current flow, but because of the resistance. Keep in mind that E=IR, and that when I gets very small because R is very very large, E does not go to zero.
If the VCE of a cut-off transistor did not increase then there would be no use for transistors in switching circuits. Consider Kirchoff's voltage law!

 

This thread keeps repeating that the datasheet is accurate but contradicts itself? No.
The written spec's show ranges of Vce sat Vbe when saturated and leakage current when cutoff.
There are ranges because they cannot make a certain transistor part number all the same, some are better than typical, some are worse than typical and many are typical which is an average spec. The graphs usually only show the typical.

Like I said earlier, A BC337 has a max Vce sat of 0.7V when its collector current is 500mA and its base current is 50mA. For the same currents, the graph shows a typical Vce sat of only 0.18V which is much better than the max Vce sat. There are BC337 transistors that have a Vce sat even lower than 0.18V.

You think there is a contradiction when the saturated transistor must have its Vce sat less than its Vbe sat when the Vce sat is as high as 0.7V, but a weak transistor with a high Vce sat voltage has an even higher Vbe sat voltage. The max Vbe sat for a BC337 at a collector current of 300mA is 1.2V. At 500mA its Vbe sat will be higher, maybe 1.5V.

 

The magic of exponents works with money as well. A penny today at 1.5% annual interest compounded will be enough to retire on in 12,290 years -- if you have the patience. Douglas Adams was wrong -- you too can be rich before the end of the universe.