For the last journal article I published, in a journal run by Springer, we submitted a nicely typeset LaTeX article using the template they provided.
Our best theory is that they then copy/pasted the result into Word. It came back to us as a "pre-print" PDF with tons of errors, font changes, etc. We had no good way to compare it to the original other than printing them both out and going through, line by line, looking for discrepancies. There were dozens and dozens of problems. And it was a 70 page article!
After sending them corrections we got to repeat this process 3 times.
And now they charge folks for our article.
Fuck. :(
But seriously, what you're basically saying here is that in fact it would have been better to have published it yourself - except you may not have had the same audience who would have known about the paper?
Or did they also go through rounds of peer review and ask for clarification or find errors?
The reason researchers publish in these journals is it gives them street cred. There are several open access journals - but scanty participation since their impact factor is unknown and does little to promote the researchers career.
Speaking as someone who has published myself, the paywalls are highway robbery. But, imho - it is a problem only the researchers can solve by concertedly supporting open access journals and getting involved in them (maybe a yc research idea???). The established professors would need to take the lead on this - not the underlings.
Personally - I've seen the "light" and stayed back in industry as opposed to moving into academia as I had hoped. The rat race for funding & publications & street cred seemed a lot worse there.