We preserve in general all UIDs in migrations. Fixing image references is usually done as a fixup step where you iterate over all content types with RichText fields and parse the HTML content using some HTML parser and rewrite the image URLs as needed.
not that i know of... but the two tools mentioned in the above comment should help
sure you can, but iiuc this will not have a huge benefit: or to the old-style urls (image_mini) do not work in plone5.2 anymore?
@@images/image/foo will get transformed to a filename containing a hash (automatic cache invalidation when image-content or crop region changes) by an output transform. you won't have this benefit when doing rewrites from image_mini to @@images/image/mini