Hi and thanks for your detailed reply.Yeah, the filter suggestions were a bit of a red herring. As you noted yourself earlier, they don't work on SuperNATURAL sounds, but even on the other (traditionally sampled) piano sounds, pretty much any modern sampled piano sound will have multiple velocity layers. In the old days, a piano note might have only had a single sample, and the sound designer could use filter cutoff combined with velocity envelope adjustment to help give it some velocity-controlled dynamics, and then opening that filter (or altering the velocity responsiveness of the filter) could add some brightness to the less-than-high-velocity strikes, but these days, keyboards almost always use different samples for different velocity levels. So then you might be able to use a filter to reduce it from its originally recorded brightness, but you can't use it in the direction you'd want, it's not going to bring out brightness in the sample that wasn't there to begin with (or as you put it, it's effectively already at maximum). And resonance creates artificial peaks that are unlikely to be helpful here.
Piano is typically the hardest sound to get sounding good through an amp. It is pretty much the "go to" sound to determine how good an amp is. An amp with more coloration or weaker highs or lows won't be noticed as much on those other instruments as they are on piano. Also, for most sounds, you don't have to worry about stereo vs. mono phase issues because most sounds are mono (absent any effects).
Out of curiosity, I don't think it's been mentioned, what amp are you using?
I think sometimes "better" pianos can sound worse than lesser ones. One reason could be a difference in stereo-to-mono compatibility. I also have an "uncanny valley" theory here... that when a piano sound gets closer to real, its flaws through sub-standard amplification become more apparent.
Since you're okay with the sound at home, as you say, it may well be because of the amp and/or going from stereo to mono, even if other keyboards have been okay in that environment, and so that might be the perspective from which to address the issue, rather than adjustments to the keyboard or adding a second keyboard. But on the second board topic...
I didn't like the piano playability of the EW-425.I don't know how much of it was the particular piano sound, vs. how it played from that action. My Casio CT-S500 is pretty decent, though, if 61 keys is enough and you want something small/light and low-cost. Again, no 5-pin MIDI, so you won't be able to hardwire it directly with the BK9.
But getting back to sparing you from having to bring a second board... amp is one variable to address as mentioned... and we've mentioned stereo vs. mono, but not how to address it. Short of actually running in stereo, here's something else you can try. If you're using an output from your keyboard that sums the stereo to mono, try defeating that by just using an actual mono signal that is only one side of the stereo signal. So if you're using the L/Mono out of the BK9, try using the R output instead.
Though also, it sounded like your initial experiment with an iPad sound was reasonably successful, so maybe some more time spent on that could also spare you from having to bring a second board.
Regarding the different samples at different velocities, I have noticed that when I play very firmly, a much brighter sound is triggered with the Piano tones. But, I do have to play them hard, and it feels uncomfortable to me to play like this constantly (I come from an organ background with no touch sensitivity at all). Even with the touch sensitivity at it's lowest, you still have to play hard to trigger that sample. If I turn off the touch sensitivity altogther and change the velocity to 91 (which seems to be the sweet spot), it sounds much better but I don't want to play with no touch response.
I'm thinking that you're right with the amp situation, and maybe more detailed piano samples require a better amp. It's an old Behringer PMX2000 that I use with one speaker. I connect the BK9 to it via a dual mono lead. So two TS jacks on either end and I go from the BK9 outputs into a stereo channel on the Behringer. I have the mode switch on the amp to mono and plug the speaker into the mono speaker output, so the Left and Right channels go into the stereo channel on the amp and are then both output to the speaker. I know this is not the best setup for sound quality, but it was good enough for me with my G1000, but like you mentioned, the same environment may not work well for other keyboards. I also noticed the G1000 had much more reverb applied by default whereas the BK9 seems minimal. But again, this may be because the G1000 has lower quality samples and they made it up with reverb. Also, I don't think the G1000 had velocity switching samples as the timbre of the tones sounded the same at whatever volume. If I could trigger the brighter sample on the BK9 that I mentioned earlier but with touch sensitivity on, I think that would be good enough.
Looked at the CT-S500 and it looks like a decend board and it sounds good on the YT video by Jeremy See, although they all do sound good on YT (well, most). I wouldn't be bothered about arranger features as the BK9 styles are great (maybe a bit busy at times, but they're easy to edit).
The iPad experiment did do ok but using this setup makes me a bit nervous in a live situation as I fear either hangups or crashes. I think trying another amp (with two speakers) may be my best bet to start with as that may be all that is needed to solve the situation.