Intel could just push back release schedule and let AMD float, kind of like what happened for years until K7 started ramping up.
Intel didn't really let AMD float so much as AMD capitalized on some opportunities. K6-II and K6-III were actually quite solid CPUs - they were quite capable of outperforming similarly clocked Intel parts at the time.
AMD systems, on the other hand, were plagued with poor chipsets and driver issues. K7 is where that changed.
AMD broke the 1.0 GHz barrier before Intel managed to, which was great marketing at the time. Intel saw that K7 was outperforming Coppermine and attempted to release a 1.13 GHz Coppermine. That Coppermine supposedly had terrible heat and stability issues.
Netburst was probably pretty far down the pipeline at this point. Throughout the early half of Netburst's life, K7 was trading blows with it. K8 was released and consistently outperformed Netburst - sometimes to huge degrees.
K8's release also prompted the release of the first Pentium 4 EEs - so called "Emergency Editions."
Intel hasn't really been guilty of being lazy since, though. They've followed their tick-tock cycle very aggressively. I can't see AMD catching a break very easily now.