For those who have plenty of time to spare:
https://www.imperial.ac.uk/media/im...-College-COVID19-NPI-modelling-16-03-2020.pdf
Modeling is an interesting academic exercise. But the more variables, the more assumptions and the more inaccurate the model. The question is what assumptions does the modeling make. A key assumption is how well the over 70's adhere to the rules. You can do some very basic math to get a rough idea of the impact. 20 million over 70's. Say 60% infected and 3% mortality. Imperial College use an adherence rate of 75% which leads to something 600,000 deaths.
But if 95% of over 70's self-isolate the number becomes 120,000. And if no one takes a blind bit of notice then the number of dead will be well over 2 million.
The more religiously we follow the guidance, the fewer people die. If you're older, or know someone who is older, or know someone with vulnerability, then you can do a lot of good by encouraging them to accept the rules and remain indoors, or if they venture out, to ensure they don't come into contact with anyone else. And if you can help them on an ongoing basis to adhere to the rules even better.
Of course if everyone who has the virus isolates for a week, or two weeks if they have family at home, then in theory over time the number of infected persons should fall rapidly or at least not escalate as rapidly.
But then in theory, if all movement outside and external social contact was prevented and enforced by law for a two week period (with very limited exceptions), the epidemic could be over in a few weeks. Which is more or less what China did by welding people into their own homes etc.