You misinterpret on at least two counts, but you bring up yet another good point
First the good point. If you just learn random things in general you will never use 90% of it. I agree, that would be a great time waster.
One cannot learn a real language in a week, though. Python or Perl can be learned enough for a small script in a week, yes, but beyond trivial subroutines there are stacks of advanced books in each language. Take the time to read them. Go through the examples line by line. Mastering a language will save you and your company untold hours a month because you know what can be done and, instinctively, how to approach things.
One cannot learn a technology in a short time, either. If you don't know TCP/IP or symbol tables or hashes, and you are a programmer, learn them. If you are a programmer in the coal or nuclear industry, for example, learn different aspects of it the steam cycle. Look into the marketing aspects. What do salesmen encounter? What are the competing technologies? Learn something about them.
I am not suggesting doing any of this using the company's money, either. If you work 12 hours and get paid for 8, then make sure you put in an honest 8 hours work.
I believe there is a strong payback for both the company and the workers if the workers have kept up. Companies show they agree with this in actual practice because it is not the worker who has kept his nose to the grindstone all 12 hours that gets the rewards. It is the one who has kept up and can therefore contribute more in the future.