Their existence was very public and very acknowledged during the Cold War, at least by the United States. The US did quite a few "publicity" blitzes relating to its tactical capabilities, the most famous of involved five men standing directly underneath an exploding atomic bomb high above them, as a way to illustrate that cities would not necessarily be harmed by tactical warfighting above them. Of course, the details of systems and their actual deployments were kept quite secret, but certainly by the period of the arms control treaties (detente and beyond) it was an "open secret" that the US had such weapons deployed globally and it was publicly assumed the Soviets had them deployed at the edges of their controlled territory as well.

It's hard to explain a non-event, but here are three factors that seem significant to me:

• The US/NATO and USSR both relied on them heavily for their war planning. For the US in particular, these were not just weapons that might be used in actual "lower-than-full-armageddon" conflicts, but were also key diplomatic tools used to reassure allies in the US "nuclear umbrella" that the US was committed to having their back against a conventional attack. (This aspect of US nuclear deployments is often unappreciated, but is still a factor even today.)

• The treaties that regulate nuclear deployments (like INF, SALT, START) generally regulate by delivery vehicles (e.g., missiles, subs, bombers) and rely heavily upon what are known as "national technical means" to verify compliance. This means, in essence, that the treaties work by assuming that a nation can, with satellites and overflights and so on, tell from a great distance whether the treaties are being complied with. So whatever you regulate with them needs to be detectable from a great distance. ICBM silos are super easy to see from satellites, and its easy enough to count bombers on runways and to keep track of when a nuclear-armed submarine goes in and out of port (and to count how many missile tubes it has, etc.). So these relatively large objects are pretty easy to verify, and thus good fodder for a treaty. Some categories of tactical nuclear weapons could be treated this way, but many are too small to verify this way, or look too much like other conventional arms. (Imagine trying to verify whether someone was or wasn't deploying the Davy Crockett system, of which you could fit half a dozen in a car.) This doesn't mean you can't have a treaty regulating this sort of thing, but it does create immense difficulties if you want a verifiable treaty — it means you need to create some kind of alternative inspection regime, one which you might not want (because it would involve, for example, enemy inspectors on the ground near your secret military deployments).

• The category of "tactical nuclear weapon" is itself pretty nebulous and ill-defined. It's sort of a catch-all category for things that don't fit into what became known as the "nuclear triad" (which itself, again, focused mostly on delivery vehicles), which range from the aforementioned nuclear mines and artillery to air-to-air missiles, anti-ballistic missiles, gravity bombs that could be considered tactical or not depending on the target, and so on. "Tactical" really is just contrasted with "strategic," and that's really more about what your war planning intent is than the technology itself. Now, all categories of weapons are a little arbitrary — the difference between an "intercontinental ballistic missile" and an "intermediate range ballistic missile" is an arbitrary definition of range — but "tactical"/"strategic" has always been recognized as a totally fuzzy definition, and fuzzy definitions make treaties difficult, because treaties need to be pretty precise and pretty unambiguous if you don't want either endless litigation over them, or your enemy finding some kind of loophole. So, again, delivery vehicles are a much easier thing to relegate by themselves — "you can have only X many bombers of a certain capability" is much easier to regulate than "you can only have X many weapons that adhere to the following vague list of properties" or "you can only have X many weapons assigned to this kind of target category" (the latter being impossible to verify without having access to war plans).

OK, so those are a lot of reasons! But it is worth noting that there was at least some ambition, in the post-Cold War, to explore ideas under which you could imagine regulate nonstrategic weapons. In 1997, Presidents Clinton and Yeltsin created a framework for a START III Treaty, agreement that "their experts will explore, as separate issues, possible measures relating to nuclear long range sea launched cruise missiles and tactical nuclear systems, to include appropriate confidence building and transparency measures." Well, that's all that came of that, because START III never got off the ground. And that is not exactly a treaty — it is sort of baby steps towards a possible treaty.

You could, in a very loose sense, thing of the ABM Treaty as a treaty regulating (one class of) non-strategic weapons, and the INF Treaty similarly applied to some systems that could be definitely used in tactical ways.