Some of what you say is simply untrue.
First, one of the images above features a road that is, at its maximum width, two lanes in each direction, not including turn lanes. Fully most of the major roads in Toronto are exactly that width and many are wider.
Second, completely regardless of street width or number of lanes, the stretch of road in question on this thread is, I would say, objectively as hazardous and unwelcoming to pedestrians as any of the examples in the images above, if not more so. It's simply not a question of opinion that the York-Bay-Yonge-Lower Simcoe offramp is a terrible pedestrian environment. And, as others on this thread have suggested, there doesn't seem to be much in the new plans to ameliorate that situation.
Third, it's not logical to conclude that Shanghai or Hong Kong opted to build that pedestrian infrastructure in response to existing pedestrian demand. What's more likely is that those intersections were identified as having especially poor pedestrian experiences and so were candidates for improvements—much like is being discussed on this thread for this particular intersection. Assuming that pedestrians aren't traversing an area because they just don't want to is a terrible urban design principle. In most cases, as we see similarly with inadequate cycling infrastructure, proper construction of pedestrian- or cycle-friendly infrastructure must precede higher usage of it.
In general, of course, simply copying urban design principles applied to nearly any situation in another city or country for use in our own city is a bad idea. That's not what I'm suggesting—instead, I am presenting four solutions to similar problems used in other cities. Looking to see how problems have been solved elsewhere is essential.