Do you have any examples of these theoretical constructs? I find most of the things I learned in my CS degree useful for commerical software development and can't think of anything I had to 'unlearn'.
Of course, if you're dealing with truly big data, like a FAANG company or something, this may not be true at all. But the reality is, 90% of the software needed is satisfied by a standard CRUD setup, and you will not need to use any of the algorithms learned in school. This can be frustrating for a new graduate, who feels an urge to do "real" development and get past the simple stuff. It is mostly simple stuff, and should mostly stay that way.
So basically, anything that wasn't CRUD (sort algorithms, processing that avoids just loading it all into memory, etc.) will usually be overkill. But, where you work it could always be different.