I disagree. The end user doesn't need to know the difference between 4 spaces and a tab. Tabs are just whitespace.
Tabs are uncommon but convenient whitespace. Commas are extremely common content. Tabs are a vastly better delimiter.
If you are wrapping source code in a CSV, a) you're doing it wrong, and b) you'll get bitten by newlines just as quickly! If you're including content that requires specific whitespace preservation, just escape the (usually rare) tabs.
TSV certainly is not perfect. But it solves the major problems for 95% of CSVs, and it's just as convenient for humans.
I do agree that one should not arbitrarily munge content. But note that HTML does munge whitespace, and we've never suffered meaningfully for it.