If I was going to push film, I would not start with this developer. *But* if you really want to do it, you have a couple of options.
a) You can increase the bath A time. This is similar to extending development in D23 (not exactly - Thornton's A formula is not as alkaline). The Bath B time won't matter much, so the usual 3-5 minutes should be fine. How much work the second bath will do, and what it will do with the density curve is up to you to find out.
b) Replace the B bath with sodium carbonate for single use. That is more alkaline than the 20g sodium tetraborate option, and will punch up the compensation and the effective speed for some subjects.
You do not say what you plan to photograph. If it is a low contrast scene, the I'd look at option (b). It it is a high contrast scene (stage work, perhaps), look at option (a).
You will have to run tests. This developer is unusual in that development occurs in both baths, and your controls for bath A are time and temperature, and bath B are alkalinity and temperature (with time as a minor variable). This makes for a non-linear response. I presume that you will be developing by inversion - things change a bit using rotary development.
Personally, I'd look at a different developer with some documented push behavior rather than mess around with the Thornton one. And I use the Thornton developer for most of my work at the moment