I found out a couple of days ago that the Setup portion of the ASUS Z87 Deluxe/Dual and ASUS Z87 Deluxe/Quad are byte for byte identical. The original ASUS Z87-Deluxe (3 different models) is different, though.
I am attempting to rebuild the Dual/Quad setup manually at the moment, but it is taking a while. Have to add the missing code and forms etc, as well as the other adjustments. Then I’ll have a look at Migration to the original “Alpha” setup.
Hi. I think you’re doing an excellent job! And that’s what I wanted to know: how to merge the TPM 2.0 extensions from SABERTOOTH Z87’s .CAP file into the Gryphon’s one (being both with the Z87 chipset). Here’s a thread/topic I created yesterday (https part missing):
win-raid.com/t9403f54-TPM-support-for-ASUS-Gryphon-Z.html
Hope you succeed so I can replicate your job in my situation.
Mario.
I have gone full crazy person and replaced the flashback components of my board into the sabertooth 2302. It appears to be taking. We will see if it lights on fire. I expect some serious brokenness, but if it comes up at all, with tpm2.0 I feel confident I will figure out the formula
EDIT: I thought they were close enough, but I guess not. Did flash, but no display, op code on board looks identical. gonna flash tpm to 1.2 and leave it alone for now
Don’t do what I did, it messes up your TPU SPI flash and makes your board not post. I have a ch341a on the way to hopefully recover. The problem is described here (I hope anyways) [SOLVED] Asus TPU unit firmware GUID
There are two TPU chips on the deluxe (one is on the back). I am hoping the physical switch slider actually uses the same SPI chip (was only able to find one) to feed both TPU’s.
I may be more comfortable trying a similar approach if I can recover this thing, just will take the TPU components with the usb flashback ones too 
Just an aside, these things use some of the most horrendous model/make strings I have ever seen (windows powershell, gcim win32_bios). I may end up patching those too to something decent.
Status update: Initial calculations and adjustments made to script. Net difference in strings is 11 (some strings are exclusively ST and some are exclusive to ZQ. Also some were eliminated). Next, I have to adjust the CMOS default settings, since (like between ST2103 and ST2302) it has grown by 9 bytes.
right now I am doing a recheck to make sure things look right, then I’ll begin the manual rebuild. Its taking a while, but should be worth it.
I’m still working on getting TPM 2 working on the ASUS Z87 Deluxe Quad (and dual), however, the latest Microsoft PC health checkup app indicates that now that is not the only limitation to be faced. Now it says that both the TPM and CPU are problems. For the socket, you can’t get anything newer than 4th gen. Mine is 4771. Everything else is good to go it says.