Reading through QBO papers, a recurring theme is disruptive anomalies not only observed (1987, 2016) but predicted [Osprey et al 2016] several times a century. Periods vary from 22 to 36 months in [Maruyama and Tsuneoka 1988]. This noisy data set covers not quite 70yrs. There does not seem to be any third-party validation of a Lunar-forcing hypothesis, which does not (yet) explain disruptions.
Ocean tides are far more certainly Lunar-forced. While locally disrupted by bottoms, shores, currents, storm-surges, etc., they never as a whole slip or skip beat.
We can reasonably make this prediction- That accumulating QBO statistics will either increasingly converge on Lunar statistics, with the noise events cancelling, if strong Lunar Forcing is valid; or the statistics will not converge, driven instead by independent planetary harmonics and hypothesized influences like Climate Change.
There is no doubt that QBO is slightly sensitive to Lunar tides, if not harmonically forced to high order.
Maruyama, T.; Tsuneoka, Y. (1988). "Anomalously short duration of the easterly wind phase of the QBO at 50hPa in 1987 and its relationship to an El Nino event"
Newman, P. A.; Coy, L.; Pawson, S.; Lait, L. R. (28 August 2016). "The anomalous change in the QBO in 2015–2016"
Osprey, Scott M.; Butchart, Neal; Knight, Jeff R.; Scaife, Adam A.; Hamilton, Kevin; Anstey, James A.; Schenzinger, Verena; Zhang, Chunxi (23 September 2016). "An unexpected disruption of the atmospheric quasi-biennial oscillation"