EXS is limited to 64 voices perc instance. Logic can load 64 instances in a session, so I guess your polyphony (assuming your computer is fast enough) is really 64x64.
For me, what makes the EXS version most desireable is Crossfading instruments for the performace patches. For example, the performace legato violins (piano and Forte) are one Xfade instrument. I don't know about you guys, but that's a big deal to me. The limitations of giga 2.5 are thrown out the window. EXS has no limits to zones/layers/samples/ etc, etc, etc.
Also, EXS can work with 24 bit samples. Not only that, I understand that you can have a single instrument patch that has both 16bit and 24 bit samples co-exsisting in the same patch. I only mention this because I believe it's relevent as far as future developments. I assume it will be easier at this point for VSL to press forward with EXS than giga, but who knows?
For me, what makes the EXS version most desireable is Crossfading instruments for the performace patches. For example, the performace legato violins (piano and Forte) are one Xfade instrument. I don't know about you guys, but that's a big deal to me. The limitations of giga 2.5 are thrown out the window. EXS has no limits to zones/layers/samples/ etc, etc, etc.
Also, EXS can work with 24 bit samples. Not only that, I understand that you can have a single instrument patch that has both 16bit and 24 bit samples co-exsisting in the same patch. I only mention this because I believe it's relevent as far as future developments. I assume it will be easier at this point for VSL to press forward with EXS than giga, but who knows?