Chaosmancer
Legend
On the contrary. Unless you establish a baseline arguments are meaningless. I could suggest taking fighters out of the core because I've never had a fighter at my table - but this would be an utterly ridiculous suggestion once you put things into context. Your suggestion here is nothing but special pleading. And even if it was about halflings, threads drift. Trying to be some sort of martinet saying "You will only talk about what I want to talk about and are not even allowed to put my arguments in a wider context" just makes you look as if you are aware that you do not have an argument.
Not Hussar obviously, but there is a significant deviation between Fighters and halflings.
Firstly, as was mentioned previously, most of the classes fall on a very small array of percentages. With 13 classes, if they were all exactly equally popular you would expect each class to rate 7.7% or so. I think the lowest class is 6%? The highest is 11%? That is all within a single standard deviation if I remember my math correctly, and so it can't really be stated that any class is particularly unpopular.
Additionally, we need to address narrative role. If we take halflings out of a setting, most of the time, you would never notice. Eberron you would notice and that's about it (while dark sun has unique halfling lore that ties to the beginnings of the world... it is such hidden and secret lore that taking them out doesn't actually alter the setting significantly). However, if you take a class like the fighter out of the game... you no longer have non-magical soldiers. A massive part of the fiction just vanishes.
So, I think recognizing differing realities between races and classes is rather important. I wouldn't cite raw percentages though as the biggest factor behind halflings needing reworked though, I would cite the ease at which they can be made irrelevant in every official setting barring one.
And as I've been saying throughout it would be ridiculous to cut halflings without cutting the less popular and less thematically identifiable gnomes. And it would be ridiculous to cut both the small races at the same time. The reason half-orcs can be cut fairly easily is that they are not only unpopular but there are other candidates that can rival them for their niche - and indeed are doing so without being in the PHB.
I still will never understand how people think that gnomes are not "thematically identifiable" they have two very powerful and deeply rooted themes, and most of their personality traits are identical to the halflings. And I think their lack of popularity comes from three major sources in the 2017 and 2019 data.
1) Intelligence is their primary score and it is one of the weakest if not the weakest scores in the game.
2) Intelligence was only useful for Wizards, and elf, human and Tielfing wizards are insanely popular. And unlike the Lightfoot halfling which has some rather nice abilities to make them better rogues than average, the Gnomes lack anything to make them better wizards than humans, high elves and tieflings.
3) The majorly thematic class for gnomes is Artificers. And they are relatively new. In fact, if you don't count the Eberron setting book, it is perfectly fair to say that it just got released with Tasha's. Meaning a lot of tables haven't had a "gnome class" like they have for most of the other PHB races.
I truly think it is not the lore and themes of Gnomes that have docked their statistical popularity, but their mechanics. Because Gnomes appear in far more fantasy literature than halflings ever have.