Geekbench is programmed differently than a standard app. It doesn't have the typical idle time that a standard app uses AND it loads the CPU to the maximum at the same time. That creates the big peaks/valleys in power draw that iOS is now looking to smooth out in order to prevent an auto shutdown. So if you run Geekbench, all you're establishing is how much throttling Geekbench itself is creating at that particular voltage level. You're not actually measuring whether or not a standard application would be throttled at all, and that's what the concern actually is: people running standard apps that have normal idle time built in and aren't designed to max out the CPU to test it's theoretical limits. Do those apps really experience throttling at nominal voltage? Geekbench can't tell you.
Again, you keep sticking to this idea that the whole thing is a Geekbench issue. I hear you. Let's assume that that is absolutely right. Is Geekbench a valid measurement for speed of a device? For years and years- and every single iDevice launch- one key measure of how much faster the new is vs. the old has been Geekbench. Where were you and/or Apple poo-poohing Geekbench as a valid measuring tool in all those countless articles & posts of the past? Why is Geekbench only wrong now... here... used in this one way that doesn't look so flattering toward Apple?
And again, Apple has CHANGED something in response to a Geekbench measurement: replacement batteries are now $29 instead of $79+. So apparently Apple respects what Geekbench results are implying to do something about it. Is Apple wrong here? Why cut the price of batteries and offer the service of replacing them if the measuring stick is as wrong as you keep trying to imply?
But again (for the third time now): if it is as you say, the old Geekbench measurement was throttled too. So when the iDevice was new and those reviews were written, THOSE Geekbench numbers- as impressive as they were vs. past iDevices- were also throttled down because of how Geekbench is "programmed differently" and how iOS handles those differences. Let's allow that idea to be the truth for a moment...
Those scores readily discovered in the public domain (reviews) establish the speed of a given iDevice when it's battery was new. That's X.
Now upwards of years later, iDevices feel slower and Geekbench scores imply that Apple throttles older iDevices with iOS upgrades... causing Apple to actually admit that "YES WE DO" throttle them, using the aging battery as the excuse... or rationale, depending on one's point of view.
Do what Apple is offering: spend $29 and have them put in a brand new Apple battery. Run Geekbench and the older iDevice should be back towards X... unless there is other code in iOS that also throttles older iDevices (not related to power management).
In short: whether one accepts that Geekbench is a flawed measure or not, it's flawed approach would apply now (after a new battery is installed) and then (when the original battery was new). Now's X should be fairly close to then's X after Apple puts in the new battery... UNLESS the battery spin is mostly spin and there is other code in iOS that is there to slow down older iDevices. We'll all see soon enough.
If replacing the battery completely gets older iDevices back toward the original X, those who feel that their older iDevices are a lot slower than they were when they were new should notice a very positive difference in speed. And if battery replacement doesn't really do much in that direction- that is, it still feels relatively slow- flawed or not flawed Geekbench tests are going to show that the original X was still much better than the present day X, implying that there is more to this than just battery power management code.
