Last year, there were about 300,000 books published–just in America. According to BookScan, only about 200 books per year reach a 100,000 copies sold. That means you have, statistically, about a 0.0007% chance of selling one-tenth of a million, much less a million.
This is just bad math. Books written 10 years ago may have finally reached the 100,000 mark. How many books were published 10 years ago? Your 300,000 number also includes re-editions, which generally doesn’t entail much, if any, extra effort on the author’s part. Your numbers are also just for print books, not ebooks. Some of those 300,000 books were also authored by the same person and, in fact, a person who writes more than one book can sell 100,000 copies between all of his or her books without any single one of them selling that many. Finally, even if the logical element of your argument weren’t completely bad, there is also the arithmetical hiccup that 200/300,000 = 0.0007, which is 0.07%.
This is not to say you’re wrong in your point that it’s hard to get rich off of selling a book, but it’s not nearly as infeasible as you’re making it out to be. Regardless, the only surefire way to never sell a book is to never write one, and if you have supreme confidence in your writing skills, go ahead and try.
I personally think it doesn’t matter why someone wants to write a book. If they’re disappointed with the result they get, or if they produce something mediocre, oh well. Maybe it can be an object learning experience, and the next time they try, the results will be better.
I am always in favor of choosing production over consumption, and I’d even go as far as to call it socially irresponsible to discourage attempts at creative production. (Especially with bad math)