It's also known that the body's response is based on light (known to impact melatonin production), and that light treatment in the day might ameliorate whatever effect the dim light or darkness has.
The lights we're talking about can, at best, light a small area around them to a brightness that's 10% daylight. The most extreme lamp he mentioned does 30%, again in a small area. Hardly a lighthouse.
It seems natural to me that if you hypothesize that lack of sunlight contributes to sleepiness or depression in winter months, then you'd want to treat that with something approximating sunlight as best you can, or 100,000 lux.
So we don't know for sure whether this works or how effective it is, but there are good theories behind it to test. Let me put it this way: suppose I put a lamp with 100,000 lumens in your bedroom and activated it shortly before dawn or, just for fun, at midnight. Would you wake up? It's clear that light does something to wake you up, and that it's harder to sleep in bright sunlight in darkness. Well, winter is short and dark for many people.