If you can point out where my goalposts moved I’d be happy to entertain the idea, but I’m not seeing it.
To clarify for you, “moving the goalposts” means that I changed my definitions over time. You’ll notice that in my comments I never changed my definition of what it means to have good battery life. I know sometimes turns of phrase are easy to misuse so I hope that helps you out.
There’s no winner or loser here. We’re just discussing technology. I’d appreciate if you tried to add conversation value rather than just dissing me personally.
Panther Lake is an impressive chip. The only MacBook Pro that can achieve 20+ hours of battery life at all with any setting is the 16” model that comes with the largest battery capacity allowed on a commercial airplane. It’s really not framework’s achievement, it’s the chip that’s so good, and that’s great for consumers because you can find a lot of competition on the market that has the coveted “all day battery life” without compromising on performance.
Panther Lake is impressive for x86, but decidedly middling when stacked up against a field including ARM.
The problem is that x86 can (almost) match ARM battery life but at much worse performance, or match the performance at much worse battery life. That's what the statistics don't show you when they show either max performance or max battery life tests.
I repeat it rather often on this site but it's because they're such impressive numbers: an M1 Ultra gets 70% of the performance of an RTX 3090 at 25% of the power envelope, and it beats out a desktop class Core i9 from that era at 30% of the power envelope.
Nvidia, Intel and AMD would commit murder for those kind of perf per watt numbers.