If it’s constantly 5-7 years behind isn’t it moving at the same speed just with an offset?
If it's a constant offset like you say, and given academia publishes and industry often doesn't, that might suggest it's dependent on the rate of advancement in academic research. Not in this field so I may be wildly off.
only if they are always going in the same direction
That may depend on the field. My experience (in bioinformatics method development) is that people in the industry can't afford to work on state-of-the-art problems. But once a problem has become established and it's important enough to be worth their time, they will eventually come up with a better solution due to their superior resources.
Business moves incredibly fast; academia, not so much.