Strange... IC13 combines /SLT33 with Z80 /BUSACK signal, and I don't know why (normally /BUSREQ and -ACK aren't used in an MSX). 
Seems to have something to do with that speed control IC, maybe it's interfering with normal /SLT33 function. Thoughts:
- Maybe the DRAM's /CAS (not the generic /CAS from S1985 btw!) is really as good as /SLT33 in this case. Luck really is on your side, dude!
- Maybe signal on pin 3/4 of IC13 is what's needed here (as opposed to /SLT33 signal on pin 2).
Yeh, pin 2 of IC13 doesn't work. Does the above mean try pin 3 or 4 of IC13 or 3 & 4 combined? 
I have S1985 datasheet thx (since, ehm.. 20 years or so)
If nothing else works, I guess it's /CE back to the DRAM's /CAS (=IC13 pin 6) like you had before. Can't figure out why /SLT33 (or IC13 pin 3/4) wouldn't work though. 
OK mate, will do. Hope it still works 
Connected to pin 6 on IC13.
now fails second pass page 7
Second time fails second pass page 20 & 23
Seems to slow down and speed up at times on the second pass.
Used to be same speed all the way through and passed every time
Only difference the schematic shows, is a ferrite bead (FB2) in between. So pick up in same point as before (IC16/17 pin 16).
If that works as before, perhaps best method would be to pick up that IC13 pin 3/4 signal through a series resistor (47 to 100 Ohm or so). Yeah sometimes it's tiny things like this that make difference between fail/succeed. 
Bl**dy hell, still failing TestRAM
Update: failed 1st 2 x tests. Just turned off, and tried again, now passed, strange
Nope, failed again. Just the odd page, every now and again, on the second pass
I don't get it. Everything connected the same again as what worked okay before, and now it doesn't?
Check your work: no short circuits anywhere? Damaged pcb traces? Some other wire came loose? Compare current connections with pictures you posted.
