It's probably two things: (1) institutional inertia and (2) actual value of FICO 9.
(1) Slow adoption is normal.because big banks are usually very slow to change to any new model (even if it is really good). The danger in shifting to some new unproven model is significant, so they usually take years before they put a new model into production, i.e. using it as the formal decision maker in actual CC or loan granting. But during that time they are often running the new model in parallel to see how well it predicts risk compared to the one they are used to. It took years for big banks to shift to FICO 8 (until last year Wells Fargo, for example, was using a FICO model released in 1998) and FICO 9 won't be any different.
(2) FICO 9 may just simply not have much in it that helps lenders and CC issuers -- FICO 8 may be basically just as good. The new technologies that conceivably could have been added to it are not there: e.g. the use of trended data. (In contrast, Vantage 4 is using TD, from what people say.)
A helpful analogy might be Microsoft operating systems. When institutions perceived no value in Windows Vista in 2006 (they were happy using Windows XP) they simply waited until the next big version of Windows came along that they liked (Windows 7). Then again the same thing happened with Windows 8: institutions were happy using Win 7 and saw no value in Win 8, so they simply waited it out until Windows 10 was released. Many financial institutions may be be perceiving FICO 9 is just not having much value and they plan to wait till FICO 10, leapfrogging over 9 just as people leapfrogged over Vista and Win 8.
