Observational results for bubble motion at high frequencies in field-accessed devices have been reported in the literature. The predictions of a model for bubble propagation under T-bar tracks are compared with these results for a frequency of 100 kHz, initially at midrange values of drive and bias fields. This comparison is helpful in refining certain features of the model so as to bring its assumptions nearer to the physical situation. The matching between predicted and observed results for the variation of bubble position with drive field angle is now much improved. The refined model is then used to calculate results at other values of drive and bias fields for which experimental data is available. The agreement between theory and practice is shown to be quite close except when the bias field is high.