This question may not be particularly answerable.
The only available information about this specific case is in the original article (Shankar et al. Ann Indian Acad Neurol. 2012 Jul-Sep; 15(3): 221–223. doi: 10.4103/0972-2327.99728); they say only:
The anatomical site of viral latency and the mechanism of reactivation of rabies, a neurotropic virus, remains an enigma to be unraveled.
Other reports in the literature of long incubation periods (>7 years, the previously reported maximum seems to be 14 years) are scattered/old/speculative, e.g. Gavrila et al. Ann. Inst. Pasteur 1967 Vol.112 No.4 pp.504-15:
Discussing these cases of prolonged incubation periods, the authors point out that in rabies the incubation period can vary within very wide limits and that it is possible in areas where the disease is enzootic that, in very exceptional cases, the disease can develop after an incubation period which may sometimes greatly exceed the maximal limits now admitted, and that, as for example in cats [this Bulletin, 1965, v. 62, (1106)], stress conditions and so forth may lead to the generalization of the virus until then quiescent at the point of entry.
Plotkin 2000 Clinical Inf Dis 30:1 says
the reasons for this long latency are unknown ...
in the absence of specific prophylaxis, the virus becomes hidden, only to invade the CNS weeks, months, or years later. Where is the virus during this period? There is some evidence for slow replication in muscle cells and for latency in macrophages, but there is no certainty that these account for the long incubation periods.
There has been an experimental study on skunks (Charlton et al. Acta Neuropathologica June 1997, Volume 94, Issue 1, pp 73–77) that suggests there is some infection of muscle cells at the injection site, but otherwise not much is known about the mechanism.
I would guess that vaccination would be effective if it were administered soon enough, but it will probably be hard in these surprising cases to identify the disease soon enough.