You make a statement that X number of people die from the flu in a year. This number is where flu is put down as a primary cause of death. i.e. this does NOT tend to include someone who was in hospital with the flu who had a stroke/heart-attack. (As you so strongly point out with a 70% estimate the US is defined to be unhealthy). Taking some leanway on this we'll be optimitic and say only 67% of the people are unhealthy at time of death from flu. This means we can triple the likely number of flu deaths if every death involving the flu was counted toward this. (pre-existing cases likely not included in this metric)

This gives an estimate of 150k flu deaths for 2017-18: https://www.cdc.gov/flu/about/burden/2017-2018.htm

Total covid deaths are 900k, so 450k/yr average.

Given covid is a novel (new) virus that is attacking a population with (zero) innate immunity or pre-existing mechanisms to cope with this new virus. (i.e. every 60yr old didn't encounter this in the first 30 years of their life and have some pre-existing natural way of coping with this) we expect the death toll to be even higher, even if it were just as deadly as the flu. The fact this is a novel virus also means it will spread through the population much quicker given there is by construction and definition zero host-based mechanisms to slow it's spread.

We will call this metric F as we don't have a good handle on what this number is, other than it's probabilistically >1.

An F in this case of =3 matches the 2 sets of data we have. i.e. 150k die normally, 450k died now.

The big question is this, which of the following statements are more likely to be correct:

1) Covid is 3x more deadly than the worst of the flu! (With an in-built, and wrong, assumption that this is not due to having something to do with the population not having an in-built immunity of some form.)

2) Covid is a new virus, be careful because this can be deadly! (This has an in-built assumption that covid is deadly to some people... and it is).

Now we can again use the data to answer this question, which is more correct. The data says that mostly at-risk people are dying. This confirms more of statement 2 that covid is killing people, but it raises a concern about statement 1).

Naturally things like the flu kills the young and the old at about the same sort of rate due to their immune systems. There is no real mechanism which explains why covid doesn't kill the young so much other than it doesn't. Because flu deaths in the young in developed world are so low due to treatment we don't have good statistics to compare between young deaths from covid and the flu other than to say they don't appear to be significantly worse. (This borne by the covid stats/testing of young patients)

Unless there is a mechanism discovered to say why covid doesn't impact the young so much we can assume it follows the general age profile of being a transmissible seasonally variant contagion.

Here is the 2nd problem.

If covid is more deadly, then by how much? This would be <=3.

The final amount covid is more deadly is some factor G.

The way G would/could* enter is that F+G explains the increased number of deaths in covid years compared to flu years.

i.e. F+G<=3 (covid could be, but is very unlikely to be less deadly, so we'll take F and G to be positive)

If the population had some normal natural immunity then G=3 by construction which means covid is 3x more deadly. This is bad, but frankly still not bodies piling in the streets bad.

We know F has some likely non-zero value and my bet is a lot more research is needed to explain it, but I doubt it's close to 0 as that implies that there is no such thing as herd/natural immunity which doesn't match reality.

Conservatively I think we can now say covid is likely between 1.5 and 2.5 times less likely assuming that the chance of the true value being somewhere in this range.

This obviously is different to decisions made around healthcare availability and reliability and correct approach to treat, but I'm going to pretend in both cases that the flu and covid have been optimally treated for the sake of argument. This is incredibly generous to your argument as it implies additional deaths can't be explained by bad healthcare availability or decisions. Unfortunately given some of the poor decisions in treating covid I think this alone could being 3 down to 2.5 by explaining, oops, we should have been treating them this way after 2 years experience.

Anyway, I don't regard 3x as being a huge multiplicative factor, which isn't to say 3x more people dying is a good thing. It's not, but obviously informed steps can and should have been taken to reduce this.

*(Could, enter this way, the unknown above is now linearly separated to first order into 2 unknowns, this is ignoring higher order effects for the sake of clarity and explanation here)

Even taking this at the worst case scenario for my case of you being wrong (50k was the true number of deaths with the flu), we end up with (F+G) < 9 with the same leniency, covid is likely between 6 and 8 times worse than 2017/18, but frankly given we don't tend to shut down society based on the numbers this is only really a concern in a healthcare setting and not massively at a population level. Esp given it's at the same level of people dying from heart attacks at a worst-case spin. We don't shut down mcdonalds because of that though...

apples to oranges gets you 6<G<8, apples to apples gets 1<G<3.

So it pays to decide on your priors carefully and given there has never been a count of "number of sick people who had flu at time of death" we can't get more accurate at resolving this.