With the massive settlement with QCOM, Apple's given themselves 6 years to try and develop competing chips which Intel wasn't capable of after many years and multi billion R&D investments. Given they're going to continue using Intel's existing 2,200 staff it's not clear what Apple can bring to the table other than a bigger R&D budget, perhaps they can eek out an edge with their tighter h/w + s/w integration. But as they're taking over the development they must think they can or at least use it as a bargaining chip for more favorable licensing from QCOM.
From a company health and security perspective it makes sense for a company as large as Apple to try and not be beholden to an external company for core technology, so from that perspective given the investment is a rounding error for Apple they might as well continue trying to compete.
Apple can probably bring much better high-level management than Intel. Intel haven't managed to make a success of anything except x86 processors, whereas Apple do well in markets they target more often than not...
While it would represent a novel venture for Apple, they could also clobber Qualcom with the threat of competing as an OEM, threaten to not only vertically integrate, but also to license their work to everyone else, for cheap. I admit it's unlikely they would actually turn OEM though.
They don't even have to be. The truth is that there is not so few 3rd party modems out there, it'd just nobody other than Chinese used them before because of Qualcomm's threat.
Even if they will use that portfolio purely defensively, it will be kind of useful for them
From a company health and security perspective it makes sense for a company as large as Apple to try and not be beholden to an external company for core technology, so from that perspective given the investment is a rounding error for Apple they might as well continue trying to compete.