In a free and open market I would predict a fairly rapid transition for Intel from fab-heavy to fab-lite and on to fabless. One thing that acts as a boat anchor on that process is what to do about the wafer fabs and thousands of people employed in them, In Intel's case those plants are all around the world. Such considerations can turn a fairly obvious direction of travel into death by a thousand cuts over the course of a decade or more.
Look to the example of IBM, which remained a manufacturer of silicon and owned wafer fabs for far longer than made commercial sense for a big iron company that had turned into a consultancy and services company.
In the case of AMD it may be remembered that the company found a fast solution to its manufacturing "problem" in the form of the sovereign wealth fund of Abu Dhabi that wanted to get the Gulf state into semiconductors and bought out AMD's manufacturing interests to form Globalfoundries.
The other obvious inhibition on Intel abandoning chip manufacturing is US government policy (see US talks to Intel, TSMC about building local foundry fabs). It is notable that China's leading foundry – Semiconductor Manufacturing International Corp. (SMIC) – is not at the leading-edge in manufacturing but Chinese government support and a strong ecosystem of has helped that company draw closer to it.
Intel may yet take the US government's dollar to continue to dabble in chip manufacturing for the sake of US government policy on strategic necessities. Swan portraying an Intel that is becoming ambivalent about ownership of leading-edge semiconductor manufacturing may have the side-effect of helping him to negotiate with a US government that will be required to pay substantially for R&D and capex.
But R&D and capex alone does not secure leadership at the leading-edge edge of semiconductor manufacturing. It also takes the vision and a desire to lead, sustained over many years. Right now, Intel seems to be lacking both.
Related links and articles: