I'd say this is newsworthy only due to Halo's history.
Back when Halo 1 was launched, it was one of the first games to use dynamic octree subdivision on height maps combined with indoor portals, which meant it was superior to the Unreal Engine w.r.t. large landscapes with caves at the time. By now, of course, all other game engines can render similar landscapes. And that's why the original USP of the Halo engine is now moot.
Cost of engine dev is too great.
When I see game credits I am always surprised how little core developers there are compared to how many other types of developers.
I imagine the real reason is that using a standard tool:
1) Hiring, it is really hard to hire people who can do core game dev engine even if you don't need a lot of them 2) Less synchronization required between teams 3) Less work being blocked by system programmers or lack of good tools 4) Lower churn of people leaving because of bad tools
I think that two additional factors are easier hiring and a shorter work-in period if you use a generic game engine instead of a custom build one
I guess it’s also not worth it anymore. Engines seem to be quite good nowadays. I guess shifting the budget to assets makes a bigger difference.
I wonder what if any other games use the Halo engines? Surely it wasn't just for Halo?