These files always come in pairs, i.e. '.W2V + .W3V' for 'waveforms' files, '.W2V + .W3V' for 'voices' files, and '.W2A + .W3A' for 'all' files. The first of the files contains the various sound parameters and it is the file that you should normally open in this software. The second contains the waveform data and is opened automatically when you open the first file (so both needs to be present in the same directory, although it is possible to open only the second waveform data file in case you have somehow lost the first file).
The 'waveforms' file contains 'Yamaha waveforms', which is basically a key and velocity to waveform map, much like an instrument in Awave Studio, but with the restrictions that overlapping regions is not allowed (i.e. velocity-splits are ok, but multi-layering of sounds is not), and that there is no 'articulation data' (envelopes, lfo's et c). You use these Yamaha waveforms as the basic building block when creating voices on the synth. When opening one of these files, each waveform therein will show up as an instrument. When saving, you can save an entire collection, or a single instrument, layer or region and 'Yamaha waveforms' will be created from these. However, for instruments which use multi-layering, only non-overlapping regions will be used. Regions in lower numbered layers have priority over regions in higher numbered layers, and any colliding regions in the higher numbered regions are simply skipped. You will get a warning message if one or more regions have to be skipped.
The 'voices' file contains one or more 'Yamaha voices' which basically correspond to the instruments in Awave Studio. These include articulation data. Unlike the previously described file type, you should be able to load up a voice directly and play it on the synth. When you load one of these files, the voices will of course appear as instruments and similarly when writing them, each instrument will become a Yamaha voice. A number of Yamaha waveforms will also be created as necessary and stored in the file - each voice can use up to 4 'elements' which each use one Yamaha waveform. This means that the maximum 'depth' of multi-layering of sounds is 4.
The 'all' files can contain many different types of data; voices, waveforms, performances, sequences et c - but only the waveforms and voices are read.
Finally, how do you move these files to and from a Yamaha Motif synth? The answer is either on a 'SmartMedia card' (you also need a SmartMedia reader for your PC), or on an USB storage device (e.g. USB hard disk, or USB flash memory). Write the files to on your PC; then plug it into your synth then and use the 'File' menu to open them.