I can think of some possibilities not mentioned in the article:
1) The longer a therapy is around, the wider it is known. That means more people have access to it before official therapy. You can go online and find CBT theory, techniques, thought distortions and how to fix them, worksheets, and other tools. This raises the chance that someone already knows CBT and has tried at least some of its methods before seeking professional help. So if the therapist then does more CBT, it looks less effective measured from the start of therapy, because the client already did some of that stuff and got whatever benefit they got from it earlier. In this case, CBT only has a high rate of helpfulness for people who really need guidance and/or advanced techniques that don't work well alone.
2) CBT is terrific at treating certain types of problems, but mediocre or useless for others. If you have bad tape, this is a go-to therapy for fixing that, and you should definitely try it. Same with any other logical or practical problem. It's also ideal for people who do better with facts, logic, numbers, or other objective things than with subjective things. But if you are feeling unheard, your emotions are bent, unacknowledged memories are gumming up your subconscious, or your biochemistry is out of whack, then CBT is not ideal for those problems and won't help much. It is possible that certain types of problem are more or less common in different decades. If the problems presenting now are something other than logical/practical ones, this therapy will seem less useful overall.
Bottom line: If you have head problems that you need help with, start by identifying them as best you can. Then look at the available options for treatment. Each type of treatment is good at some things and bad at others. Pick one that's a good match for your problem(s). Try it for a while. If it doesn't help, drop it and try something else.