I'd argue the whole point of productivity gains is that they do outpace pay. The idea is the same work generates more value. Some of that extra value can be passed back to the employee, but if all of it is passed back to the employee, then the goods produced don't actually get cheaper. If nothing gets cheaper, there's no incentive for a business to invest in tech that makes employees more productive.
The data in the piece has a lot of problems with it too, but I think the core assumption is fundamentally off-base.