Even at the low end of the general prediction ranges (200k), coronavirus would come in as the third leading cause of death for a normal year. So it will, alone, boost the numbers by a fairly good amount even if a lot of the people it kills would have died from their other illnesses.
While car accidents will go down, I would expect suicides, murder-suicides of families, etc. to all go up by a significant amount due to Covid-19. When people are allowed to go outside again, we’re liable to start discovering a lot of doors that are still locked and need to be broken down.
Cabin fever ain’t nothing.
Lack of light, poor nutrition, insufficient exercise, etc. will contribute to reduced health for many. Constrained hospital resources will cause there to be insufficient investigation of health issues for many. We should expect more deaths through general poor health and poor medical care.
Likewise, insufficient money due to lack of income could cause more cases of thefts even though those homes have inhabitants. This may lead to more murders, one way or the other. Alternatively, those people in poverty may end up homeless and, subsequently, stay that way. That won’t kill them this year, but it will shorten their lives considerably and eventually factor into things.
It’s plausible that shelter-at-home could kill more people than letting people go out and get Covid-19, particularly if you intelligently divided things and told those under 50 with healthy lungs and no underlying conditions to keep living life and those over 50 or with other complications to shelter-at-home varying to complete quarantine, depending on their risk factor. Being shut-in isn’t healthy and for some it may be more unhealthy than an extra-strong flu.
You would need to analyze a lot of research across a wide variety of sources to try and predict which is better, overall. But, I guess, in this case we’ll find out through direct experimentation.