I've been advised I need to look at bivariate outliers in my twin dataset and that it used to be done with %p in old Mx. Is there an equivalent function in OpenMx?
Yes, but you have to change the fit function somewhat, using vector=TRUE in the mxFIMLObjective(). This returns the vector of likelihoods, which you have to log and sum and multiply by -2 in order to use instead. However, the full functionality of mx%p is not yet implemented, and would need to use the number of observed variables for each data vector (which could be extracted from the dataframe). It is something that could be done with a helper function, as the formulae are not very difficult to express in R.