Most of the calculations are for the nook simple touch and by association would apply to the kobo touch and kindle touch since they use similar/the same processor, IR touch emitters/processors and eink pearl screen.
The bit about backlit is in general refering to full tablets like the ipad, fire, nook color and nook tab.
An eink reader (obviously varies some) is going to burn somewhere roughly 100mw while "on" and reading averaged out. A tablet is going to burn roughly 1.2-3w while on and reading depending on the screen brightness (supposing nothing is running in the background). Eink based ereaders have batteries ranging from I think around 3-4wh (I think about what the new Kindle 4th gen uses) up to around 5-6wh (5.5 for the nook simple touch, not sure on the kindle touch, but probably 5-6wh). Those tablets use batteries ranging from roughly 12wh up to about 26wh (the 7in tabs use a bit less power because of the smaller screen, but their batteries also tend to be around 12-16wh and the bigger 9.7-10.1 tabs use more power because of the bigger screen, but also have the larger 22-26wh batteries).
So despite basically a cell phone capacity battery (5-6wh) eink readers use radically less power (as I mentioned, roughly 100mw). Since I don't really know exactly how the processor, screen renderer, memory and other systems are operated during a touch event and afterward, I couldn't tell you if everything stays "powered up" to a higher power state for a prolonged period of time, or if it is short. However, without any real testing, based on pure observation during normal use, it seems like screen refreshes take roughly half a second under normal reading, but if I hit the screen by accident (or on purpose) or a paging button within at least 1-2 seconds of the previous page turn, the action causes a page turn much faster which suggests everything remains powered up and doesn't need render a new page (just draw it to the screen) or power up the eink screen (if there is even a state other than "drawing" and "off") and what have you. Longer than that, there is some cut-off where screen redraws/refreshes take longer, but I don't know where that happens. More than 15-20s and it takes that half second to redraw (instead of maybe 200-300ms that an immediate follow on touch seems to take), 4-5s? Not sure.
So all that being said, I really don't know exactly how much power a page turn will result in, overall. I can tell you based on expected usage case and amount of time BN and Amazon claim for their devices, and empirical evidence seems to suggest that average power consumption during relatively normal reading is about 100mw. My guess is at around 30pph reading rate, about 15-30% of the power is used by page turns and the higher power state that it incurrs and 70-85% of it is used by system idle.
So if you read faster, the amount of power used by page turns and the resulting higher system power state immediately following a page turn will be higher, however, at a "normal" reading rate, the "doing nothing but on" state likely consumes a large majority of the power used by the nook ST; however, that is still a pretty tiny amount of power being used.