Compared directly, My HD58X is modified to have lower bass response and thus is very similar to S2 in all aspects, shows that the HD660S2 sounds slightly 'more pleasant' to me. Just a bit more 'refined' in the treble. To be the HD650 has better treble quality than orig HD660S and (modified) HD58X but the HD660S2 is just slightly better.
Don't have the original HD660S though so cannot compare directly but could compare to the same HD58X
I still have and had at the same time.
To give you an idea how my HD58X is below my modified + 6kHz passive filtered HD58X
(which is not
the same as stock HD58X) v.s. HD660S2
Below stock HD58X
(Romania produced) and HD660S2
do sound different and to me (may not be so for others) and HD660S2 wins.
NOT looking at the price difference though.
I can't look into the wallet of others nor say whether or not the (small) increase/difference in SQ is worth it to someone.
The extra € 400.- one has to pay for the HD660S2 thus may not be worth it to most people but could be for someone that appreciates the difference.
Would I trade-in my modified HD58X for a HD660S2 ? Nope as I already have a HD800 which is 'better' in aspects I value the most.
Do I think the HD660S2 is worth € 600.- ? Not to me. Do I think it sounds really good ? Yes.
While I got the FR pretty much the same on FR plots in ideal circumstances and pinna-less there still are differences in the actual drivers.
This is evident when looking at the seal measurements (the HD660S2 reacts much better) due to differences in damping scheme.
Not only impedance (and thus reaction to say the HP output of an AVR) but also damping, resonance frequency and possibly even modal break-up differences in the membrane that could, certainly with many different pinnae, result in audible differences IRL situations.
So similar in FR (modified HD58X) but not the same.