Rather, I think houses should be able to ride out flooding without significant damage, like the stilt houses above. In fact, the stilt houses are really overkill- just a couple of feet higher should probably suffice for most houses in affected areas in Canada. This can be done over time through municipal and provincial laws, and selective insurance coverage. The overall goal would be to minimize damage, not avert it completely.
Not practical in most situations. You may save the house, but you have sewer/septic/water intake; hydro polls, cable/fibre, roads, and important public buildings like schools/fire halls and hospitals.
To make every one of these flood-resistant across a large area is going to easily crest 10B and possibly a good deal more.
Flood resistance is practical only on a limited scale, typically in peripheral to flood risk areas, where your protecting against the 100-year plus storm with just a bit of extra caution.
The only other reason to use this strategy is for a very limited area, deemed high value/essential to remain in its current location, but that's a pretty high bar to meet.
The solution? Don't build on a floodplain, or even close to one. What you posit is a solution to existing homes to get them above the flood level, but what about all the essential services and ability to function as an island in inclement weather?
Entirely correct.
Now the asterisk is how we define the floodplain? Do we use the 25-year storm?, The 100-year storm? The 200-year storm? etc.
I typically think the 100-year storm is the correct bar; but we need both updated flood maps that correctly show flood risks as they are; and reasonable projections of where the risk will be 100 years out, given climate change.