Some of the people I know got COVID and they loss the smell. After discussing with them about it, they told me that they only smell one persistent smell, everywhere (like onion). Is this frequent for people with loss of smell ? How this is possible ?
At this time any hypothesis regarding the parosmia seems to be untested. Speculative sentences in lay and technical sources point toward the olfactory bulb. (I have seen indications regarding the cilia of the olfactory neurons, the bulb itself, and the blood supply, but this seems to stray from our topic)
When we look in more detail, the olfactory bulb consists largely of glomeruli (balls of neuropil about 0.1-0.2 mm in size), which receive input from specific kinds of olfactory neurons that receive from specific odorant receptors (ORs). See this figure for the situation in mice. The dorsal region of the olfactory bulb is mostly one type of these, Class I ORs, and the ventral part is all Class II. (more) If you are following this research you might want to try to figure out if the odors people report are mostly Class I (more hydrophilic) or Class II (more hydrophobic), which might allow you to hypothesize something about the relative strengths of signals from these regions in the model you develop. Reading anecdotes is scientifically useless, but fascinating - this man describes citrus as tasting "extremely oxidized", which certainly fuels a speculation along that line.
Answering the question in any good way would seem to demand good human olfactory tissue from people who catch Covid, suffer ongoing parosmia, die from something else, and then donate their brains to science. I'm not sure we're going to see a definitive answer that way soon. It would be worth keeping an eye out for fMRI data. I suspect that well-focused high-frequency terahertz might be used to image or even stimulate olfactory glomeruli (phones with 'smell-o-vision') but for now that's science fiction.