Let's say you are fitting a seasonal ARIMA model with regressors with some daily data and your best model candidate is: ARIMA(1,0,0)(2,0,0)[7]. By 'best' I mean a model that simultaneously deals with autocorrelations in the residuals and forecasts reasonably.
Now, according to this source, this should not be a good model:
Rule 13: If the autocorrelation at the seasonal period is positive, consider adding an SAR term to the model. If the autocorrelation at the seasonal period is negative, consider adding an SMA term to the model. Do not mix SAR and SMA terms in the same model, and avoid using more than one of either kind.
Does having a second seasonal AR or MA term really invalidates your model; which assumptions are violated?