I discuss that possibility in the post. It seems more plausible as an explanation for linear growth in confirmed COVID-19 cases than for linear growth in deaths attributed to COVID-19. For the latter, the situation would have to be something like… there’s no tendency to use tests on the most serious cases (many of whom die), and instead they test a fixed number chosen at random of those who come to medical attention (of whom the most serious are a small, fixed fraction), and only attribute a death to COVID-19 if it was of someone who tested positive.
A constant level of testing, leading to a constant number of confirmed cases some of which die?
I discuss that possibility in the post. It seems more plausible as an explanation for linear growth in confirmed COVID-19 cases than for linear growth in deaths attributed to COVID-19. For the latter, the situation would have to be something like… there’s no tendency to use tests on the most serious cases (many of whom die), and instead they test a fixed number chosen at random of those who come to medical attention (of whom the most serious are a small, fixed fraction), and only attribute a death to COVID-19 if it was of someone who tested positive.