I agree. As someone else mentioned, during the open playtest, many of us were sending in comments addressing this (and similar issues), Despite our best efforts we were ignored by the designers.
I guess the question becomes, is the fix worth the effort?
Changing language proficiency or tool proficiency is one of the easiest possible switches in the game. The mechanical impact is nearly zero, only blipping up past that if your DM is one who is really into using languages.
And, the Rogue knowing Thieves Cant makes perfect sense in the PHB, standard Assassins, Thieves and Arcane Thieves (called tricksters to harken back to the prestige class) would all likely have connections to enough criminal elements either in their past or in their present that the proficiency makes sense.
The cases when this is not so might be considered edge cases.
Look for a similar problem, you have racial languages. Why do all Tieflings know Infernal? I can chalk it up to Hell Magicks and call it a day, but it is a serious element you have to consider while world-building.
And, talking about edge cases, what about the Dwarf/Gnome/Elf/Halfling (how many people even remember they have their own language now?) who is raised by humans and has no contact with their people, having been led to believe they were human or some other race this entire time. For some reason people love this concept, but then the question needs to be asked, why do they know their racial language, it makes almost no sense for them to know it.
Or, how about a Dwarven soldier who held a rank in the army, fought in the goblin wars for 20 hundred years, never learned the language of his enemies? That doesn't make a lot of sense, but there is no way for a Dwarf Fighter Soldier to learn any languages other than Dwarvish and Common per the rules.
So, you either come up with an explanation that works for you, or you homebrew it or you switch out the languages.
It is
an issue sure, but not one that needs a comprehensive or even official fix.