I am a big supporter and advocate of opensource software. But after so many years of both using and contributing to the opensource movement, I've realised that the popularity of an opensource software isn't really a reflection of its code quality. Minix 3 and FreeBSD are better than Linux. TextPattern is / was better coded than Wordpress. OpenSSL was insecure for so many years till its fork LibreSSL.
Inexperience of the early founder(s) of the project is one factor for poor code quality (that often does improve as the founder becomes more experienced and / or other experienced devlopers contribute to the project). Documentation and testing is another.
But another factor is money which dictates the design of the open source software.
Mozilla is a good example of this - despite the millions of dollars they have received, they deliberately avoid modularising their software. Gecko, the web engine, should be available as a stand alone piece of software that anybody should be able to use to create another browser or some other application with it. But there is no easy way to do this because Gecko has been tighty coupled to the browser code. And this is very intentional - Mozilla makes it money from browsers, and hence despite working on an open source project, their code is deliberately convulated to make it harder for others to create another browser from it to compete with their product. (Yes, you can fork their existing mess, but it's a sub-par product because of the way it is coded - you would have a really tough time creating a native UI for it, for example - and so you just end up with an imitation of their browser).
(And it is not as if they can't do it - as they realised they were losing the browser wars on mobile, they've finally released the Gecko engine for Android - https://mozilla.github.io/geckoview/ - hoping others would be motivated to create Gecko mobile browsers and they continue to remain relevant on the mobile platform).
SQLite is another example where the code is opensource, with the most liberal source license, but the testing code is closed and unavailable. Again a deliberate choice made to hurt any competing product created with their code, i.e. a choice dictated by commercial interests.
Both the above examples highlight how the open source spirit is somewhat crippled and sacrificed when a lot of money comes into the picture.
