It is a falacy that you need to test all combinations.
Even if you do not, you still need one group for each drug, plus 1 placebo, plus in the U.S. you may need male and female and different racial types. Some drugs have a measurably different effect depending on sex and race. I wouldn't know about COVID-19 drugs. You might need an obese group, age groups, or some other groups. I wouldn't know.
Even with thousands of critically ill patients you end up with many different small studies. The effort is fragmented. Also, those thousands of patients are not available for long. Most of them get better in a few weeks, or die.