The marginal difference of starting a new coronavirus chain

What is the marginal difference you make if you start a new coronavirus chain? Say that you infect 2 people who both infect 2 people who infect 2 people etc. Many people in that chain would have gotten the virus anyway, but probably some wouldn’t have, so you might think you’re causing the overall number of coronavirus cases to increase. But if the coronavirus continues until enough people have had it and we get herd immunity to it, then is the marginal difference you make to the number of infected basically zero? You would just change the particular people who get it. Or do you make a small difference in expectation? This could be because the herd immunity threshold is vague and there is a chance that the people in the chain that you started who wouldn’t have gotten the virus otherwise won’t make a difference to whether we’ve reached the threshold or not so they’re extra infected people.

But even if you wouldn’t increase the number of infected by starting a new chain, you might still increase the number of deaths due to the virus, because you would speed up its spread and thus increase the number of people who have it simultaneously. That would cause overcrowding of hospitals and more people would die because they wouldn’t get the care they need.

Here are some reasons for why we should take good care to not accidentally start a new coronavirus chain:

1. Slowing down. By not infecting new people you slow down the virus and thus help reduce the overcrowding of hospitals (and thus save more lives). But by slowing down its spread now you might make the governments less likely to place harsh measures to stop its spread, so you might increase the number of cases at its peak time when it matters more. Thus, slowing it down early might increase the number of simultaneous infections later and thus cause more deaths due to overcrowded hospitals (thanks Tomi Francis for this last point).
2. Partial identity reasons. By starting a new chain you would endanger people you spend a lot of time with (friends, family, partners, co-workers). You might think it’s permissible to be partial towards people close to you and thus you might care about the identities of the infected people. By taking extra care you might spare people close to you even if you don’t decrease the overall number of infections.
3. Non-partial identity reasons. There might be non-partial reasons for why it would be worse if the people close to you are infected instead of other people, for example if they’re in the risk groups (e.g. have asthma) and thus more likely to die/need hospital care, if they’re doctors/nurses needed for taking care of sick people, if they play some other vital role in society that cannot be put on a hold during a pandemic or if they do otherwise very valuable work and their deaths would deprive humanity of that work.
4. Rule-based reasons. If everyone takes extra care, then we might be able to stop the coronavirus from spreading and actually decrease the number of infected. Rule-consequentialism, deontology and possibly also evidential decision theory would then recommend taking extra care to not spread the virus.
5. A small probability of a large payoff. There is a very small probability that the coronavirus would not be contained because of the chain you started (while it otherwise would have been contained). Given that millions of people could die from the virus, act-utilitarianism combined with expected utility maximizing might recommend taking extra care because of the possibility that you tip the balance between containment and a global pandemic. But the probability of this happening might be so small that even with the high utility number the EU of acts that reduce infections (staying home etc) might be negative due to extra hassle (act-utilitarianism + EU might still accept reason 1.) mentioned above because of the much higher risk of overcrowding hospitals. It would also accept some non-partial identity reasons mentioned in 3.).
6. Reducing the number of infections. You might make a marginal difference to the number of infected people by starting a new chain (?)

Is the claim that you wouldn’t make a marginal difference to the number of infected plausible (while still possibly making a difference to the number of deaths)? Do you have any thoughts/objections/further points? (I left out some caveats to keep this post shorter. Also, I only discussed short-term reasons to avoid spreading the virus.)