Wondering why a layered PSD file with adjustments, when converted to a smart object, and then reopened, the smart object, the arboard within the smart object is larger? is this something to do with a bounding box? Trying to answer a bunch of designers.
Certain operations require to expand the area they operate on to avoid edges being cut of or dirtied up with "garbage" such as layer styles or filters that dilate/ extend/ expand/ generate additional pixels and may have artifacts at document borders. It works as intended for your own safety.
I guess but once you make the smart object, if you want to open and use as an un-layerd file the canvas is bigger...