"The chart below shows that, of the top cryptocurrencies by market capitalization, some - like EOS, Cardano, Monero, and Bitcoin - have had a lot more commits than others - like Populous or Bitconnect". Since Bitconnect is a now defunkt ponzi scheme and not a cryptocurrency, their lack of Github activity makes sense :) This analysis would be more helpful if the author properly curated the projects / tokens to include in the study.

Is this only looking at one repository per project? Some projects, like Ethereum, have a number of repositories:

https://github.com/ethereum

Here's a site that tracks exactly this:

https://cryptomiso.com/

Seems like there would be a big difference in metrics here if committers squash their commits or not.

I love this. I've been kicking around the idea of building an automated crypto index buyer that orchestrates purchases/transfers on exchanges, around metrics like commits as well as use metrics like the ratio of on-chain value transfers to speculative on-exchange trading.

Would be nice to couple this with a Linux distribution that made it easy to create an offline wallet for new cryptocurrencies that become interesting, so you could immediately shift your coins off of exchanges.

A lot of them will probably become worth very little/go to zero, but there's a reasonable chance that the Amazon, Apple, Microsoft, Google etc. of the crypto world already exists as a fledgeling project.

commit metric would be so easy to fake though

Coingecko has been tracking this too, for example https://www.coingecko.com/en/coins/ethereum/developer

Another factor to consider is the content of each commit. Is it just copying from another project or making trivial changes, or is there real work being done?

I was curious about market cap to twitter followers or posts in https://bitcointalk.org.