There's nothing wrong with a failed startup, but this was a poor copy of miners and mining chip producers with much less funding and in some cases bootstrapped to profitability. The only thing that set 21 apart was an amazing ability to razzle-dazzle VCs.
After failing to mine Bitcoin profitably themselves, 21 came out of stealth mode with a staggeringly stupid product: mining chips for toasters. The idea was that you would buy a toaster (or lightbulb or something) with one of 21's mining chips and it would waste energy while mining miniscule amounts of Bitcoin. This was ridiculous on its face, since the amount of Bitcoin that could be mined on these chips was far less than you could simply buy for the cost of the electricity (not to mention the chip itself). Also, this amount was too small to even be transferred, given Bitcoin's transaction fees!
I'm not exaggerating, this product was never, and could never be viable. Still Balaji was able to razzle-dazzle the tech press into trumpeting it as some sort of innovation.
Once this failed, they re-branded as Earn.com and started making a grab-bag of Bitcoin hackathon projects, such as paying you Bitcoins for reading spam, or answering quizzes. This was always an extremely uninspiring and ultimately non-viable product, but at least it wasn't complete vaporware.
Then Coinbase, a very successful and competent company, threw away $100 million to help 21's VCs recoup their failed investment. The only explanation that seems rational is that Balaji had some very compromising footage of Brian Armstrong at the a16z holiday party.
I'm not sure about his past career (some biotech thing), but Balaji's legacy in crypto is to waste enormous amounts of money on blatantly ridiculous boondoggles. It's unfortunate that Coinbase wasted so much time and money perpetuating this.