The springs definitely change over time but not as much as you’d think in some examples.

The goal is to get the spring prototypes to sound close to what an original Model F would have sounded like fresh off the production line, not an exact match to what they sound like today where corrosion and humidity have definitely affected the sound of the springs (more so for the poorly stored Model F’s in humid-damp places than for many of the F’s in my collection).

I have a home movie of myself typing (for the first time!) on the family’s Model F XT keyboard from our IBM PC.  Back then these keyboards were much newer.  I have reviewed the sound of our Model F.  (We no longer have the computer or keyboard though).  It did not sound much different to many of the Model F keyboards today – still with a characteristic buckling upon actuation and a small amount of reverb.  The reverb is in line with some of my best condition Model F keyboards today.  No two Model F keyboards sound exactly alike; I am aiming for my favorite original Model F, a 1984 F122.  With age the springs should even more closely approach the older springs.

Some of my earlier forum postings on the springs:
Updates on the spring research – samples matched original XT springs!
Over the past week I reviewed the new spring samples as well as all the old samples.
In summary I was able to match almost perfectly a spring sample to an original IBM spring pulled from an XT! I classify the archetypical spring as having some middle-pitched ringing typical of a Model F, but not too much (characteristic of a more well-used Model F keyboard or one with significant spring discoloration/corrosion). The original Model F springs have a wide range of sound – many in their current state are more higher pitched than what I’d like.
The key is to use Japanese wire (the right material and quality of wire are both important) and to let the springs age for at least one year (don’t worry I won’t be holding up the project to age the springs!). Samples I rejected in Dec. 2016 now have that perfect sound between loud ringing and no ringing – who would have known! Not even the American wire held up to the Japanese metal in terms of sound. The Chinese wire had the highest pitch, followed by the American wire (still too high), and then the Japanese wire which matched the ringing pitch.
I have compared many original XT springs from different keyboards and determined that the ones that ringed the most tended to be corroded/dulled but not rusted reddish brown. So I expect the springs will emit a bit more ringing (what I prefer to call the sound after a key is released – not including the plate reverberation sound) in the years to come. For now I expect the new Model F keyboards to have a mild ring matching an XT stored properly and with little to no corrosion. I do prefer the louder-than the average Model F ringing but it will have to develop after heavy usage and time!
I also performed some accelerated corrosion/rust tests on some of the sample springs. Generally a bit of discoloration/corrosion on the springs is not an issue for sound or performance, but reddish rust seems to add some tizz/squeaking to the springs in some cases, like with original XT springs. The production springs will have a protective finish.
Next step is to get Japanese wire spring samples made with the current supplier and with the updated/confirmed material choice from the lab. Key molds, inner foam, and inner assembly plates are still in production and expected to finish in a few weeks so we are not bottlenecking anything by getting things right with the springs.

I am hoping to loan out review units once the early bird batch arrives from China, but please keep in mind that every Model F keyboard in my own experience has a unique sound and feel, meaning a reviewer’s sample Model F has little chance of matching the sound of another original Model F or of one of these reproductions (the repro’s should sound similar to each other given they are all new and from the same early bird manufacturing batch). Given that, many a reviewer will certainly say something like “The sound is not the same as my original Model F.” Even the thickness of the inner foam and its state (disintegrating older foam or replacement newer foam) plays a role in the reverberation (thicker foam in my experience makes the keyboard quieter and less reverberant/musical). Two AT’s in my collection sound different from each other and from an XT. Two of my F77s sound different from each other and other F’s. My F107s sound different than my F122s. Even one of my 1985 F122s sounds different from my 6110344 from mid-1984. And it’s not just the sound – the state and type of springs also plays a factor in the typing feel/experience – a long time Model F user may prefer the worn down springs in his original F because of the reduced actuation force, compared to a new F (my new F’s take some time to break the springs in so that they are even smoother). Maybe a reviewer prefers quieter keyboards or keyboards at a $100 price point, which would rule out a new Model F.