About AS80: it needs the -x switch to compile z80 code 
Alexey, I still have a request:
As you may have noticed, there are a few programs that use a 2nd PSG to have more music channels, those programs are mainly written for the MFRSCC.
Since it's impossible to detect the presence of a PSG at address 10H (because the device is write only) they check if there is the string "MFRSD" at address 4010H.
Now, it's possible to change the C2 firmware to have the FPGA PSG at address 10H, but unfortunately the programs don't recognize it and have to be forced or manually patched.
So I've hex edited the BOOTCMFC.bin to have the "MFRSD" string in place, and with this trick VGMplay works, but there are compatibility problems with Victor's USAS & Zanax-Ex enhanced versions.
So I propose to have a new STANDARD identifier string that can be unique to a C2 with a modded address.
For example instead of "CMFCCFRC" something like "CMFCCFRC2", so that with a patched firmware and patched bootblock current and future programs will be able to detect it.
What do you think?
Report that bug to VGMPlay please 
. There should be absolutely no need for it to fake being an MFRSD, and I prefer it not to do so (I’d rather things be what they say they are).
The only thing needed for the detection is for the PSG to be readable at port 12H. That is the preferable way, and this is how e.g. GR8NET and homebrew internal 2nd (real) PSGs are detected. MFRSD detection is a fall-back specific for this hardware. But even when it’s write-only, VGMPlay still attempts to send data to it, so if it doesn't play there must be another problem. Though once I’m sure there are no write-only PSGs out there anymore except for the MFRSD, I’d like to remove that undetected usage.
As for the C2 having its own unique ID in ROM, that of course never seems like a bad idea. If there’s absolutely no way for the C2 to expose PSG read access, I could decide to check for it in VGMPlay. But I prefer a readable PSG rather than an exceptional check.
