Well the entire video game industry is predicated on the concept that increased financial risk results in increased financial reward. Hence why consoles have long employed the "razors and blades" business model of moving systems at or below cost to make up the difference on game sales. Also evident in the ever escalating cost of AAA game development where bigger = better.
Same in the movie industry where production costs have risen steadily to fund major summer blockbusters that dominate the entire cinematic landscape.
Most media produced for mass consumption works on the concept that the more money you put into something the higher the profits on the other end will be, assuming it is a good product.
Nintendo clearly doesn't believe that or at least believes that isn't a viable strategy for them. They're probably right, but that trickles down to how they roll out products with a much more conservative eye towards initial manufactured numbers, total manufacturing capacity, and how quickly they want to meet demand in exchange for risking overproducing as demand dries up.
You're trying to match anecdotal evidence with quantifiable anticipated sales, that simply isn't a valid strategy for a company to go by. Even when companies market test products they assume a pretty healthy decline from people who say "yes, I'll buy that" to the number of people who would actually show up to buy it because there is a proven trend of this happening (same with people saying they'll vote, as another example of inherent dishonesty in polling). Nintendo miscalculated the NES Mini's demand but likely did so with sound reasoning at the core, just sound reasoning built around very conservative ideology.