No, I do not use ARDL. The ARDL model is useful for viewing cointegration and can also see the effect.
While to test the causality can be used bivariate causality and multivariate causality. In a simultaneous model, causality can be tested by estimating predicted values and calculated reidual.
It is a good econometric model to estimate the long run relationship between variables. However, applying unit root test guides you using ARDL or another methods.
All other co-integration methods(eg. Johansen co-integration) requires your variables to be of same order for their application(either all I(0) or I(1)).ARDL method is best to use when your variables are a mixture of I(0) and I(1).It helps you find out long run relationship between variables. Further using ECM you can also estimate short run relationship. Hope it helps.