CPU/GPU chip design, which Johny Srouji brought to Apple, is far different than novel modern digital communications theory and design. Especially when considering Qualcomms 150,000+ patents covering the field they essentially brought to life decades ago, which will be difficult to not step on.
Not saying it's impossible. Just extremely unlikely. It's a field where fractions of a dB in performance hugely matter.
Much better for Apple to purchase chips (or possibly a design Apple can fab) from Qualcomm and know you're getting the absolute best performance. And saving a lot of money in the process.
My last engineering job was at a small company (eventually acquired by a huge US semiconductor company) that developed full-custom communications oriented signal processing ASICs for the cellular telecom industry and military/govt. customers. At one point we considered developing some special communications chips that could have overlapped some of Qualcomm's tech in some aspects. Or overlapped some Qualcomm tech we weren't even aware of. It didn't take a lot of thought knowing that going forward would have been very foolish as they could have easily litigated us to bankruptcy.