From the datasheet table 6.5 the power dissipation is limited to 985mW. Whereas from table 6.4, the R_th,j-a is 68.3K/W which at a power dissipation of 985mW yields a rise of less than 70 Celsius. I can understand that for a general specification where the ambient could be high, the power dissipation would be limited to 985mW. However, if I could ensure a low ambient (laboratory use only) and/or a lower thermal resistance (airflow), is it reasonable to assume that the driver can dissipate more power?
For instance, if the ambient was always 30 Celsius or lower, would it be alright to dissipate (150 - 30)/68.3 = 1.76W?