Elephant in the room
Bitcoin will have to give something to make it a global currency. It will probably need to come from several sources. Interestingly, Bitcoin will require a hard fork at some point in the future due to quantum computing, regardless of policy and opinion. Quantum computing will always create a requirement for different types of encryption, as it will override SHA-256. It is used by Bitcoin among many other security protocols that it will make obsolete around the web. The community should be given plenty of warning before this becomes a major threat model for Bitcoin, among other issues. Interestingly and unfortunately, I assume that the type of hard fork to be used according to a quantum-resistant hash algorithm during such an event will be an extremely boring and uninteresting fork.
Given the controversial nature of such an event, if there was a time to run a hard fork from the Bitcoin network, it would basically be around the time required by mathematical law. And during this time a lot can be done to improve Bitcoin not only in terms of scalability but also in terms of other features; for example increasing Bitcoin's privacy.
Unfortunately, due to a problem based on game theory known as the median electoral theorem, even if it benefits the smallest number of people in the community, the most likely event to occur is the least significant. In fact, this kind of problem is so prevalent in politics that the name, the median voter theorem, exists in almost all political structures, from small-scale operations in homeowners associations to federal governing bodies.