I could write a much longer response, but the main problem is that the human body is very complex, we understand virtually nothing about the operation of the body on a molecular scale, and the current science is hilariously primitive. One excellent example is this paper [0] which may look fine on the surface, but is in fact deeply flawed. This is the unfortunate standard of "research" today.
I will summarize it: The European Food Safety Authority was asked about the safety of added formaldehyde in animal feed. They Googled and found that your body generates lots of formaldehyde during the course of an average day, 878-1310 mg/kg bw per day or ~60g/day for an average person. They compared that number (big) to the amount of formaldehyde in food (small) and concluded that it's fine.
Of course, that's not how the human body works! That formaldehyde is produced in cellular pathways that are specifically designed to synthesize it and break it down. In normal cellular pathways, this is mainly the formaldehyde dehydrogenases, along with occasionally catalase and other things. But when you eat or breathe formaldehyde, it follows an entirely different path in the body and interacts with countless things on the way. Take the ~60g/day that a human generates themselves and put that in the air in a room. You will have burning eyes and trouble breathing at a minimum; if the room is small enough or you're unlucky you may die!
The core idea here (and this is something that I have written about for some time) is that to truly understand something, we must understand all of its pathways. We cannot estimate from raw masses alone. And yet, almost all research today is based on crude estimates and not detailed investigations of pathways and their effects. This occurs at all levels, big and small, from prestigious studies to meaningless filler papers. Take the recent scientific history of vitamin D. We have known for decades that higher blood levels of vitamin D are correlated with better health in virtually every single way. Mental health, gut health, heart disease, cancer, eyesight, you name it. The correlation is so powerful and so obvious that it is one of the most notorious in health science. If you just supplemented vitamin D you could cure the nation, no? Massive amounts of money were invested in this, including a fifty million dollar longitudinal study. It all came back as having virtually no effect [1]. Because nobody thought about the pathways! Even in the US, where most people are sun-deprived and where our food is often fortified with vitamin D, over 90% still comes from sunlight. Sunlight has many pathways in the body, and increased vitamin D levels are just one of those pathways. Swallowing one element of a many-element pathway cannot simulate the human synthesis of vitamin D! This should be obvious! Your car might not run because of bad needle bearings, but the solution is not to add them to the gas tank! First, you must always understand.
Taking this back to formaldehyde and your questions, I like the questions, but we cannot know the true answers without knowing the pathways. "What levels can a human accept without harm" could vary wildly; maybe there's a priming response, or it varies by time of day, or it's nearly harmless when pure but dangerous with a common impurity, or ten million other factors, which is why so much research doesn't replicate, and why medicines have such inconsistent results. Perform research as it is done today (test human or animal exposure, get numbers, data, do not seek to understand) and you will get results as we have today: huge volumes of data, declining health, and a lack of understanding.
Oh, and if you're curious about the EU and formaldehyde, since that paper is a decade old: Plenty of papers continued to demonstrate formaldehyde's "safety" in low amounts, the EU continued to permit using formaldehyde in animal feed for years afterwards; eventually, enough workers got sick from handling the formaldehyde feed that they banned its use anyway.
Very informative reply, thank you. I'm busy and want to reread it. Will respond, may take a day or so.