Elderberry has some ability to bind ACE2 and another coronavirus site. It looks good so far, is effective for flu of course, but like most of these things only studied in small trials with self-reported outcomes.
The best evidence I've found is for probiotics, it takes a few months but they can double or triple flu vaccine seroprotection rates in healthy elderly people, and probably reduce the incidence of seasonal colds and flus.
These drugs are very specific for HCV and don't disrupt host processes to any meaningful extent. But different drug combos are specific for genotype 1 and genotype 3.
How could this work for a completely different virus?
Hep C drugs need to reduce viral load to "undetectable" because stealth features of HCV interaction with immune system mean it is a persistent infection which immune cells have little chance of clearing.
Drugs only need to reduce COVID-19 by a decent amount - the immune system is always working on it and can clear it eventually, the drug just needs to give meaningful help (enough so the immune system needn't over-react would be cool).
I have other questions about pharmakokinetics - it's easy to get drugs into the liver, because that's where they're detoxified, but the lungs?
Note that rational drug design software has twice pinpointed the herbal andrographis panniculata, a traditional treatment for viral infections and cough.