@laceflower First, I have a concern that you are teaching her a stay-ish behavior while you are away, in that you are rewarding stay and not rewarding breaking the stay on her place cue. I worry that a sensitive pup could feel stressed knowing that you want her to stay in her place the whole time and having to decide to break the stay for water or to relax comfortably. Instead, I would suggest you go outside the door, say the release word and toss a treat away from the place, and shut the door. With my clients, we always combine SA work with crate training because it avoids destruction, but if you don't want to use a crate the place and release should work. You can also teach a dog to back up on cue and that could be an alternative for the door rushing. You don't want to train the dog to wait eagerly and intensely for you to come back. You want to train the dog to relax until you come back, hopefully not thinking about you much while you are gone.
So my ritual is to cue the dog to chill out because you'll be back before she can worry. You should not need to do a slow ramp up of time away every time you go. The ritual is to set up a relaxing environment. I use the crate, the music, the lighting, the chew or Kong, etc like a little spa setup. You can do whatever, you just have to do the same steps (preferably in the same order) before you go. You want the things you choose to be mostly a de-escalation of excitement levels. So you aren't making a fuss or doing a bunch of behaviors for high value rewards. But knowing that you turn on her favorite calm music or whatever you choose as your ritual every time you leave on an absence that isn't long enough to be scary will help her predict what will happen. Dogs find routines really reassuring to help them know what's going on and what to expect.
Unavoidable fails are when you have to be gone longer than she can cope with. I assumed you were confining her for those in some way so she didn't wreck your place and/or escape. If she's not that determined, a doggy play pen around the door to create a dog-free vestibule might do it. I advise teaching dogs to keep back from the door in general, but I don't know if that would hold with a SA dog if she's feeling desperate and isn't confined. Do you practice door manners when you enter or when you open doors for her to go through?
The other thing I would note is that both whining and refusal to eat are symptoms of stress and anxiety. So if I got that in a training absence session, that would be a failure and require backing up a few steps in duration difficulty. You want the dog to not have worried feelings to properly desensitize and countercondition alone time. So I suspect you are increasing the difficulty level instead of as the dog feels calm/content at 80-90% success rate, but more as the dog feels stressed and worried but not panicked. So you may be teaching her to barely tolerate your absence instead of be fine with it. So it will be hard to stretch that time any longer because basically it's already being practiced for too long, and any other stressors could easily push her over threshold, including routine disruptions. That sounds to me also like when you would see regression like you are seeing. I would not expect a dog who is calm and comfortable during sessions to regress after going a few days without. For effective counterconditioning you need the dog to start to actually like the thing because it becomes a predictor of something pleasant.
That's another advantage to combining SA work with crate work. With a crate, once you get past going into the crate and then shutting the door, you get a lot of very easy steps of separation while the person is still in sight. The dog should be used to relaxing in the crate and getting to work on a Kong before you are making it as far as the door. You are still separated because the dog can't get to you, but it feels less drastic because you are still in the same room. A stationing behavior doesn't feel as much like separation because there is no barrier at all, so you being outside the closed door is a much bigger step, and the human doesn't get as much practice responding to the dog's stress signals before they lose sight of the dog.
It could be helpful to get her to the point she can work on a chew or Kong while you open the door and shut the door repeatedly, maybe stand in the doorway so she can see you but you are outside, etc, and use her willingness to continue to work on the Kong as your indicator that she is enough under threshold to increase criteria. The longer you can make the Kong last the more practice you can do; you could turn it into a daily meal as part of your routine with some extra tasty bonus goodies to make it special. No additional reinforcers necessary!