The CYP450 group is most effective in first-pass metabolism, which is generally the oral ROA. When you go the parenteral route, CYP450 contributes, but there are other metabolic pathways that also kick in immediately. Therefore, blocking CYP3A4 would have the most effect on orally administered ketamine. Parenterally, it might slow it down a bit but the liver has other methods once a xenobiotic is in blood/serum.
The article's statement about CYP3A4 inhibition is ambiguous because it does not state whether that's true of oral or all ROAs. My guess would be oral.
In any case, IM/IV would be faster, stronger and shorter lasting because there is no delay.