That's a really good dataset (with a really clunky visualization, but hey!)
Indeed. There are typically above 3000 deaths/week due to Pneumonia and 200/week due to influenza at this time of year. We had a smallish spike in mid-late January, where for 3 consecutive weeks, deaths were above trend at nearly 4000 & 400 per week.
If you assume a 1% mortality rate for COVID, and you assume all deaths are due to pneumonia, and you assume the entire late-January spike is due to COVID, that could be hiding 300,000-400,000 US cases. (To be clear, those are all questionable assumptions! The mortality rate could easily be either lower or higher, a meaningful fraction of deaths might not show as pneumonia or influenza, and it is normal for there to be a moderate spike above trend at some point in the flu season)
That doesn't mean that there were or weren't a few hundred thousand cases in that time -- it just means that some significant number of cases could easily hide in "routine-looking" spikes in the graph.
This does mean, though, that it is unlikely that everybody who got "the really bad flu" that's been going around this year actually had COVID -- but, especially if you live in a place with lots of people who travel to China regularly (like I do!), it isn't unreasonable to wonder if that "really bad flu followed by a lower-respiratory-tract infection that tested negative for influenza" you got in late January/early February wasn't really COVID.
(Remember, though, that while the rapid influenza test has a pretty low false-positive rate, it's got about a 50% false negative rate -- so testing negative shouldn't lead you to assume it isn't influenza)