The Post asks: Is Every Memory Worth Keeping? The focus of the article isn't about perfect memory - but that's the logical end result of the kinds of tech the article discusses. Perfect memory - or perfect-enough memory (ie, limited resolution etc) will be a thing of the near future.
The constraint won't be the cost of storage (which is the way we think about it today) - in fact, I like this this article because it points out that the constraint will be search, which might be costly for mechanistic reasons (ie, our perfect-memory search engines suck), or cognitive reasons (some memories are emotionally painful).