Simply replacing 8417 with KT-88 without compensating for the lower transconductance of the EL-34/6550/KT-88 should reduce the power output of the MKVI. Given the available surplus of output from the MKVI this may or may be noticeable to you.
As I recall the MKVI runs with a B+ spec'ed at ~500V. Just a bit higher than the MKIII. If your ac mains run high at your location I suppose you might be pushing a 6550 closer to its plate voltage limit. EL-34 and KT-88 are spec'ed higher. NOTE: It is not unusual for people to claim that modern tubes do not meet their plate voltage max spec. I have not experienced this myself. My VTL 80 had a B+ of 650V and I had no EL-34 failures nor did I expect any. FYI: I bumped into Luke Manley a while back and asked him about high B+ in VTL amps--he was surprised that I thought 650V was high.
The topology of MKVI is similar to other Dynaco amplifiers. Dynaco used 7199/6AN8 to drive EL-34, 6550, KT-88, and 8417. I would expect the negative feedback ratio to be reduced in the MKVI given the lower output at the 16 Ohm tap for a given 7199 operating point when replacing the 8417 with EL-34/6550/KT-88.
Bottom line: If the MKVI really does sound "worse" with KT-88 I would look at the circuits before the KT-88s before blaming tube/transformer matching.