IMPORTANT DISCLAIMER: I am describing a possible experiment where someone with the proper knowledge and equipment might be able to test out this "enhanced sustain" function on an existing keyboard that has a simple on/off-type sustain pedal jack. If you try this, YOU DO SO SOLELY AND ENTIRELY AT YOUR OWN RISK AND RESPONSIBILITY! There is always the risk of equipment damage, electric shock, or other undesirable/dangerous outcomes when experimenting with electrical equipment. It is also possible that this experiment could void any warranty on any of the equipment that you're using.
If someone really wanted to try this "enhanced sustain" function, and they had good knowledge of hooking a keyboard up to a computer and designing a program that could analyze, by way of the MIDI connection to the computer, what notes are being played... I bet someone with this equipment and knowledge could rig an experimental system to try this out.
You would also need to know a way to have your computer program send an electrical signal to an external device like a relay or transistor.
This also assumes that the sustain pedal on your keyboard is a simple two contact, single-pole-single-throw on/off type switch. If it isn't, then this experiment WILL NOT WORK and might even damage the keyboard! If you do not know if your keyboard's sustain jack is designed for a simple 2-contact on/off-switch type sustain pedal, then do not attempt this experiment!
Basically, the computer program could analyze the notes being played on the keyboard in a loop, say, 10 times a second. Just look at the notes below the split-point of the keyboard -- or, if the split-point cannot easily be retrieved through the MIDI connection, then just set the split point in the computer program and only analyze the notes below that point. With each loop, calculate what chord is being played. And, have a variable, perhaps called PrevChord, that would store the name of whatever chord was played in the previous iteration of the loop. Initially, PrevChord could be set to "null" or zero. When the current chord is calculated, compare it to the previous chord (in variable PrevChord). If it is the same, then do nothing and go to the next iteration of the loop to analyze what is being played. But if the current chord is different than the previous chord (PrevChord), then activate a routine to send the electrical signal to the relay or transistor I mentioned above. In either case, store the name of the current chord being played in variable PrevChord so the program will know what the previous chord played was when it calculates the current chord in the next iteration of the loop.
The relay or transistor would be hooked across the terminals of the sustain jack, and the computer program would normally set it as "closed" or ON, which would be connecting the contacts of the sustain pedal and activating sustain. But when a chord change is detected, the program and computer would send an electrical signal to briefly "open" these contacts and then "close" them again, thereby making the keyboard think that the sustain pedal is lifted briefly and then re-applied.
It is important to note that the relay, transistor, or whatever circuit you design to hook to the sustain pedal jack connections would only short/connect the contacts of the jack and NOT send any voltage directly to the wires or contacts of the sustain jack.
After that, the computer program would just resume it's repeating loop of analyzing what notes are being played, and therefore what chord is being played, on the left side of the keyboard.
Having a routine to calculate what chord is being played would not be too difficult, and for a simple test like this, you could just look for the basic major, minor, 7th, and minor 7th chords and just play those chords when testing it. That would at least help determine if such a system would work and sound good. I also wouldn't worry about inversions of chords, because you only want to simulate briefly releasing the pedal when the chord changes, not the actual inversion of the chord.
As for hooking the relay or transistor to the sustain pedal jack -- you would not need to open the keyboard at all. Usually, from what I've seen, the sustain jack is just a 2-contact 1/4-inch phone jack. So, just get a 2-contact 1/4-inch phone plug, hook wires to it that hook to the relay or transistor circuit (or whatever circuit you'd design to be able to control -- by the computer program -- whether those 2 contacts are closed or open). And then plug that into the sustain jack of your keyboard. Again, it's important to note that this relay or transistor (or whatever) circuit would NOT send any voltage directly to the connectors of the sustain jack -- it would only need to short those contacts together under computer control, which, as far as the keyboard is concerned, would be the same as pressing down a sustain pedal hooked to it...
AGAIN: You do this at your own risk! And if your sustain jack connection and pedal is ANYTHING EXCEPT a simple 2-contact SPST on/off switch, then THIS WILL NOT WORK and could even damage the keyboard!
This is a little bit beyond what I am directly able to do, but it should be possible. Certainly, it is possible for a computer to read, by way of MIDI, what notes are being played on a keyboard. And it should be possible for a computer to output a signal under the control of a program running on the computer (after all, computers are hooked to various peripherals that are controlled by the computer all the time). So, for someone with the knowledge, equipment, and maybe too much time on their hands, this could be a good way to test this sustain function.