The problem is that you wouldn't need a dozen but rather 100+. Support doesn't scale well.
The notion that customer support, just like security, is just a sunken cost needs to stop.
2. It is a free product. Would you spend 7 million a year supporting a free product?
3. 99% of complaints would be of the nature "GitHub is broken. HELP!" and really be some person who did not set the origin before pushing.
That is roughly the fully-loaded cost to employ a senior software developer in much of the world. A decent but first-line tech support agent would cost a fraction of that amount almost anywhere outside the US.
2. It is a free product. Would you spend 7 million a year supporting a free product?
Presumably for the same reasons you spent far more than that to buy and continue to operate the service at all.
3. 99% of complaints would be of the nature "GitHub is broken. HELP!" and really be some person who did not set the origin before pushing.
Then 99% of the complaints will be dealt with quickly using standard replies and references to FAQs. These aren't going to be the ones attracting hundreds of comments and presumably thousands more readers on prominent online forums for software developers.
Basically the same deal as "educational licenses" for expensive software suites. It isn't a freebie at all; it's an investment into future customer acquisition through lock-in. That's how you're paying for it.
I for one pay for GitHub. I've never had a ticket that wasn't answered in under a day, but I'd be almightily annoyed with them if I submitted a ticked like this and didn't have a response within hours.
No, front-line support staff might scale that way, but organizational costs scale superlinearly with the size of the bottom of the pyramid, so even if it did, total cost would be super linear with userbase.
It's unclear if the poster had a paid for account.