I have written two .wav files (writeWave function from the tuneR package). The first one was set as Extensible=TRUE and the other one as Extensible=FALSE.
When I read these two files (in R or audacity) the samples values differ.
Is there any way to explain that?
Regular WAV files that comply with the Microsoft specification are rather limited in the number of channels, compression schemes, bit depths and sampling rates they can have. The extensible version, as the designation implies, extends these options: MS docs.
In practice things are a bit more complicated. Some programs can write non-standard compliant WAV files that can still in general be read by other programs, while still other programs can't read the less common extensible version, even if it is compliant.
If you want to find out if tuneR is doing things correctly or not you could use something like SoX to take a closer look at the file structure.