You are not alone on this one. When you “arm” a setting, it does not show the last recorded value - it shows the deault position. This makes recording more at the end of an existing take harder. So it is not just you - it is the way it is designed. (Sometimes good, sometimes bad.)
One advanced approach that came up on the March 2018 tips and tricks video was you can have multiple behaviors on a puppet, and if you do that, when you arm one behavior you still see the last recorded value of the other behavior. Its a little tricky to use in practice.
The other way is to leverage both “Position X/Y” and “Absolute X/Y” - one kicks in before scaling, the other after scaling, but if I have a recording that changes Position X and I just need to do a bit of a tweak, sometimes I record the other setting because it’s easier.
But I think a few people would like what you are asking about - to be able to see the last recorded value so you can add to the recording without losing the current positioning. It’s just not trivial to do because of the current semantics of arming properties. (And who knows what would break if they changed that now.)
Thank you! I assumed I was doing something wrong! It would be an amazing feature to add for sure, to be able to pick up right where you left off, but I have zero clue how to design and build software, and would it would take to make that happen, lol. I'll check out that video again too, I may have missed something. Thanks again for the help!