I don't think the likes of coinbase employ developers knowledgeable in maintaining a distributed consensus system. Maintaining a project like Bitcoin is vastly different to a normal software project. The best c++ developer in the world is useless at maintaining Bitcoin if he does not have a deep understanding in a vast array of other skills needed, and the people who have those skills are likely able to earn much more money working outside the Bitcoinsphere in less stressful and hostile environments.
Just look at the background of many of the people who have worked on Bitcoin. Mike Hearn was one of the most known senior Google engineers prior to his involvement in Bitcoin (I would occasionally read his blog for SEO tips, it was very popular), and also worked for a GCHQ contractor as a SIGINT operative (according to his google+ profile) before he went to on work on Bitcoin. Greg Maxwell worked for Juniper networks and worked on the BGP protocol, a decentralized protocol used by ISP's to exchange routing information for IP addresses, a critical part of the internet itself. Adam Back has been working on e-cash systems for decades, invented Bitcoin's POW implementation (as we have all heard a million times by now) and worked for David Chaum's company DigiCash back in the 90's. Finding people with the level of skill some of these people have who would most likely have to be willing to work for less money in a high stress environment would be difficult, it's not impossible, just not something that could easily be done.
I have posted previously speculation that the endgame of blockstream was to sell services related to confidential transactions AND confidential assets to banks to help with middle/back office functions of securities trading (blockstream is working on both confidential transactions and confidential assets projects). The amount of money to potentially be made for settlement services of securities trading well exceeds the market cap of bitcoin, even in its current bubble levels.
A lot of the features implemented into Bitcoin are not necessarily actively desired by the user base and/or economy, and are implemented in a very specific way so that there is excessive technical debt and broad community consensus is not obtained prior to implementing certain features. I would opine that many of the devs working on Bitcoin currently are probably overqualified. Nevertheless, major bitcoin related companies could fairly easily fund the development of Bitcoin as blockstream does currently.