Yes. From what I understand, the 2m estimate for the USA came from a relatively early version of the Imperial College model in which the virus is allowed to let rip with no social distancing or other control measures. In reality, even if government did nothing, people would themselves do some things that would have at least a small impact on the reproduction number. That number, however, served its purpose in concentrating minds (or, in some cases, what passes for a mind) and getting the imperative for control measures into the heads of policy makers.
The lower end of the 100,000-240,000 range comes from a model in which (1) lockdown is imposed across the whole of the US (which is not yet the case, even now, I understand) and (2) the health system is not overloaded (which it almost is already, in places). Any lockdown gaps or overload will increase the number.
None of the models is necessarily particularly good, because we simply do not have good data, either on the true virulence of the virus or on the the degree of spread in the population, but they are all we have to navigate this minefield.