I think you only have a chance at implementing change if you're able to locate and leverage an exploit that the viruses within have either overlooked or can't completely secure against.
If things get too far off the rails, then the fate of the political process rests in the choices of a few powerful people whether or not to do the right thing, the level of engagement and organization of the resisting populace, the strength and independence of our various institutions, and plain dumb luck. There are many modern examples of failed democracies. We shouldn't think that "the pendulum will swing the other way"; sometimes it does, but if it doesn't then it's to late for a do-over.
My thoughts on this are greatly influenced by a recent book called "On Tyranny; Twenty Lessons from the Twentieth Century" by Timothy Snyder.
We (speaking as a U.S. person) need to do away with the electoral college and first-past-the-post voting not because popular vote and a better election system are nice-to-have features, but because our current system paints a giant target on the political process we all rely on to ensure accountability. When the outcome of presidential elections depends on a few hundred or thousand voters in one or a few of Florida/Ohio/Michigan/Pennsylvania/Wisconsin, that's a systemic vulnerability.
There are active exploits in the wild; this is less a matter of instituting reforms to make things a little bit better and more a matter of adapting to new threats in order to reduce the risk of a systemic collapse.
Other exploitable vulnerabilities in our democratic process that have been weaponized to varying degrees include: Winner-take-all primary states (on the R side), superdelegates (on the D side), gerrymandering, Congressional procedural abuse (like the filibuster in the Senate, or refusing to vote on popular measures if it's opposed by a majority of the majority in the House), congressional gridlock, and sloppy campaign finance regulations.