Sharepoint is similarly entrenched at DOE when I worked with them in 2009. Any project that gets a bid contract can't say specifically that it must use Sharepoint, but the requirements will be written in such a way that Sharepoint is literally the only system that fits all of them.
While SharePoint is not the best choice for many applications, it has enough features to make it work. Since there are plenty of SharePoint experts already in the organization, it could be cheaper to just leverage them to build some SharePoint site that just barely works well enough for whatever application. It would cost a lot more to bring in some consultant that knows how to setup the proper software to do whatever new thing the project needs to do along with whatever new service contracts that need to be signed and paid for to cover the software for years. Do you jump into the new software that might be better or do you just use the same old SharePoint that you've already invested heavily in? The efficiencies of both solutions are probably intangible and it would be difficult to price either of them. Eventually someone out there will be the evangelist to convince program management that the new software really is the best choice (Jira, Confluence, Teams, Salesforce, Trello, etc).
