Redaction is not simple. What should the software do if the user's black box is drawn over an image, some text and part of a spline on a page? Modify the image and recompress it even if it's lossy?, find the text run and remove the obscured characters (what about a half covered letter?), somehow modify spline and possibly replace it with several smaller ones if the redaction is in the middle of it? All while taking into account any transforms that have been applied to the page. Also what if the black box is over part of a pdf form? Is it just the appearance that gets redacted or the form data (which is separate). Same questions if the black box is drawn over annotations. And what about other page content like embedded videos or 3d content, good luck redacting those. But if you only redact some things the user might then justifiably expect it to redact everything.

Who owns that Issue? Who allocated budget to address? It's a Team with a Leader that misses things, not some first-year doing things "wrong".

I can empathize with the feeling but this is the case in many many contexts. Reading the intent of a user without a clear mental model of software is difficult, sometimes impossible.

I know no os that changes the format of an image when a user changes the file extension, for example, which is a pretty similar situation.

I know at least Preview on MacOS has this feature now: if you try to put a black bar over text it tells you this isn’t secure and offers to redact for you using a method that both visually replaces all characters and removes the embedded text underneath It is non-destructive until saved meaning you have a window in which to make changes before the data is gone.

Let's be honest, it's an Adobe issue