First we assume that choosing to purchase jam is empirically better than choosing not to purchase jam. And we ignore the question of how the jam selection is made, when the supermarket chooses for you: what is driving the supermarket's selection of which single brand to offer? Is it price to quality ratio? Price? Partner deals on other brands? When I start a new jam company, do I aim to make a happy consumer, or a happy grocery store manager? This doesn't seem optimal for the consumer at all.
The argument is later 'bolstered' with the example of jeans; we're told that they SHOULD be "the ill-fitting sort", that “The secret to happiness is low expectations.” If you expect shitty jam, the supermarket can remove choice to satisfy your expectation. For that matter, so can the market of jam providers.
At last we get the example of pensions, where we can at least agree on the basic point that not choosing is definitely worse than choosing poorly. We're advised that employers should choose for the employees to prevent this outcome. Again we leave aside the question of how the employer selects a 'one-size-fits-all' pension plan. Does she optimize for the best pension for employees, or for the cheapest plan up front? And how does that change the behavior of pension plan companies?
Generally, choice in the market is an important driver of improving real and perceived value for consumers. Perhaps there are alternative solutions to 'choice paralysis' that don't involve sacrificing quality. Off the top of my head, expert advisors are a better solution than what the article offers. There are hundreds of wines at my local store, but I have no trouble making a choice with the guidance of the sommelier. There are thousands of health insurance options where I live (in Germany), but I'm confident that I make the right choice for my family with the guidance of my Versicherungsberater (insurance advisor). Interestingly this is a field that is difficult to automate, not least because humans trust human recommendations more than machine ones. I can imagine this author would praise the coming "advisor economy" as the solution to fears of machine-learning-driven unemployment.
In short: everyone accepts the premise that "choice paralysis" is real, and that it causes people to make sub-optimal choices. It's an enormous leap however, to get to "therefore having no choice, and expecting shitty outcomes, is the best answer."