In short
I think your images are probably underexposed, for whatever reason.
Kodak Portra 400 is famously forgiving for overexposure, but much less so for underexposure — generally giving "muddy" low-contrast results in that case. You have very low-contrast / low dynamic range images "pushed up" into a higher register.
Long version:
Check out this histogram (from your third example; the others are similar):

You can see that all the tones are actually well within the dynamic range of the file — although everything is in a high key, nothing is smashed up against the side. Arguably, these images are basically okay for exposure. The issue is that everything is low contrast and muddy.
We can use the Curves tool to stretch them out. That gives us a histogram like this:

Now, we have darker darks and the midtones in the middle. The image:

There's also a strong color cast, which we could have corrected while we were messing with curves, but didn't.
Low contrast, muddy images are exactly what to expect from underexposed Portra 400. So, I suspect that your images are actually significantly underexposed (for whatever reason), and that the lab is pushing them up to compensate as best they can. (Probably their scanning machine just does that.)
I would:
- Take a test series with a wide range of exposure (start by underexposing by several stops and go up to over-exposing). Maybe one for each camera, in the same situation. See how the results compare between the cameras, and how they look compared to what you'd expect from the meter reading.
- In the future, aim for overexposure. Maybe set your film as ISO 320 or ISO 200 in the camera (but tell the lab to develop it as normal). This is an easy way to dial in a kind of permanent exposure compensation with a film camera.
- Get to know your meter — I don't know these specific cameras (my film camera is a trusty old Pentax K1000), but I doubt they're very smart. They don't know anything about the scene, and just aim to make everything kind of middling gray. So, they can easily misinterpret large blocks of white (like the building in your third example) or light reflected off water (in the others) as reason to make the result darker than is "correct".
- Maybe find a better lab? I'm saying this more based on the color cast than the exposure/contrast, which I doubt is their fault. Even with the underexposure, I'd expect images to come back more like:

... but:
A. Maybe they have a higher-cost option where they do more correction? But,
B. I bet that if you start with more-brightly-exposed images, you won't need it. And in any case,
C. You should definitely expect better results than even the "corrected" version from Portra 400.