It really depends on what you're flying and why.
Just like pushing the propeller forward; some places teach it because they can't think outside the very small box in which they operate. Others do it for standardization. Others leave the prop back until the power is far enough retarded that they don't have to listen to the RPM wind up (my personal choice), and some don't push it up at all unless going around.
Carburetor heat isn't simply on because one is getting carburetor ice. The manufacturer doesn't have any way of knowing the conditions in which you are flying. You may be in conditions conducive to icing, or you may not.
Flying at idle power on a hot day with carburetor heat out can lead to fouled plugs, especially with a low idle and low heat. 100LL does that. That can lead to a very rough engine or engine failure on the go-around. I've experienced that. When one applies carb heat, one is enriching the mixture by virtue of decreasing air density with hotter induction air. I've removed spark plugs from airplanes doing nothing more than traffic pattern work,to find them completely glazed over...in large part due to pushing the mixture rich for the landing, and running too much carb heat. Especially on hot, high days.
What does your AFM/POH have to say on the subject? What does your company or training organization have to say? Are you teaching in an environment where you sacrifice personal technique for the sake of standardization? Are you required to match everyone else's procedure? The answers to these questions can determine your course...but bear in mind that if you do fall in step improper practices simply because everyone else is doing it, you'd best brush up on your forced landings, at some point.
Last year I worked with some individuals who were operating in a hot and high environment. The head training person was a former instructor at a respected college with a flight training program. He insisted that the mixture stay rich for ground operations, and go rich during the descend and landing. I was able to determine that his practice was leading to a lot of plug fouling and accounted for rough engines and rejected takeoffs. He wasn't familiar with clearing fouled plugs with engine and mixture operation; he wasn't familiar with proper mixture use, and he wasn't familiar with the requirement to re-adjust idle mixture any time an airplane is moved to a new base involving different elevations, temperatures, or conditions. He knew very little about his mixture.
You're asking about carb heat, I'm talking about mixture...because they're related and affect one another.
Likewise, if you're talking about a Cessna 172 or another airplane that uses a carburetor with an enrichment feature (making mixture richer at higher power settings), your use of mixture and carburetor heat can have a bearing on engine operation (or lack thereof) at high power settings (such as the go-around). If you're leaning for high density altitude (a good policy any time one has any density altitude of any significance), one needs to take the enrichment feature into account...most instructors don't seem to even be aware of it.
Discontinue carb heat early in the approach on a day highly conducive to carb ice, you may have a problem even before you reach the runway, or may have one upon arrival at the runway. Some engine/carburetor marriages are more conducive to carb ice than others.
In many light airplanes, I prefer to test carburetor heat for fifteen seconds on the downwind, just prior to my initial power reduction abeam the numbers. At that point, assuming it isn't a red-letter carb-ice day, I'll shut off the carburetor heat, make the power reduction, and carry power throughout the approach and landing.
I think many who cater to the carb-heat necessity also cater to the idle-descent school of thought...which is a poor practice in most light, piston-powered airplanes.
Bear in mind that the carburetor has a narrow range of temperature in which carb ice is likely. If one has a chance to operate an airplane with a carb air temperature gauge, one quickly learns that partial carb heat is generally in order, and only as necessary to keep the carburetor induction airflow in the proper, desired range. Without a carb temp indication, one is spitting in the wind and taking a wild guess.
Do you need it, or don't you? It's not the same on two different days, and not necessarily the same between two different airplanes.