It is split into two parts: The Bitcoin Core reference implementation, and BIPs.
As regards to the first, the members of the Bitcoin organization have the highest control over the software, but they cannot make decisions by themselves without the consent of the rest of the group.
Authority for the BIP numbering process, on the other hand, rests on two people: Like Dashjr and Kalle Alm (although the entire bitcoin-dev mailing list can have a say in the BIP, which can only be numbered with a majority consensus).
Actions like softforks are managed by this category and not the Bitcoin Core developers.
In any of that, I would never count out the
Bitcoin community running their own personal nodes, though.
If anyone were to bribe all Bitcoin Core devs and / or all BIP people into making a change to Bitcoin that's bad for it and that the rest of the network will reject, there will be a hardfork and the longer chain ('controlled' by thousands of independent community nodes) will win.
As for further development, in such a worst-case scenario, new GitHub organizations and repositories can easily be created.
All in all, yes: these people do have 'control' over the repository, but they can't meaningfully use that to go against the Bitcoin community, as it would just refuse to update their nodes / signal a bad BIP.
I'm not too involved with altcoins, but I do believe that's kind of what happened with Ethereum and Ethereum Classic.
'The developers' who had
control over the repository, the name and tons of coins (that's one thing making it a bit different from this situation here) forked their coin to remove a badly coded smart contract.
The people who called bullshit on that, just continued running the older version at first, and then continued development based on that version, under the new name 'Ethereum Classic'.
Do keep in mind that in case of ETH, there's an influential influencer (Buterin) who heavily shifted the public opinion
towards going with his malicious chain; we don't really have that in Bitcoin, gladly.