Performance-wise there is no difference between the two; the CPU needs to grab and store all these things the same way regardless.
The first one is better for debugging and for code readability. For example, if you get a NullReferenceException on line 4 in the top code then you know for sure that currentScene is null. If you get the same NullReferenceException on line 4 of the bottom code, you have a LOT more bits that could possibly be null.
If you are going to use “currentScene” again in the function, you better use the first one.
I suggest using the first way in order to make the code more readable.