The Pentium 4 was an attempt to win via marketing (look! 3.06 GHz!) and features (look! SSE2!) rather than by pure performance. Not only was Intel competing with AMD, but they were competing with themselves, avoiding moving x86 to 64 bits because they really, really wanted Itanic to succeed.
After Itanic did nothing in the market (cue crickets), Pentium 4 being outperformed by Pentium 3 and AMD, and 64 bit being dictated by AMD, Intel reacted to those swift kicks in the ass by finally releasing something actually relevant. Competition is good!
The Core, then Core 2 lines, were significant and finally showed us that Intel was taking things seriously. They brought Apple to Intel. The fact that they were based on the P6 shows that Intel can acknowledge mistakes and can go back to what works when necessary. However, Intel never stopped trying to win by marketing, as evidenced by the fact that even in to 2009, they were still selling single core CPUs. The Core and Core 2 lines only went to four cores from marketing pressure. Luckily they kept improving with the Core i lines.
The swinging pendulum has AMD now on top and Intel trying desperately to catch up. They can only compete by throwing hundreds of watts at the problem. But the historical accounting of Intel's reaction to a good walloping does a decent job of informing us what we might see now. Fun times!