What impact, if any, do the number of bootstrap/resampling iterations have on the representativeness of the distribution of the bootstrap results?

To be concrete, suppose a bootstrap with replacement was used for the following model that we fitted in class:

lm(birth_weight~maternal_smoker,data=baby)

Suppose two bootstraps were ran, one with 1,000 iterations/resamples and another with 10,000 iterations/resamples. Would the one with 1,000 or 10,000 iterations be better? Why?

Options:

10,000 iterations; This would provide a more comprehensive picture of the population.

1,000 iterations; This would run faster and give a representative distribution of the population.