If there were actual accountability, software engineers would have a much better lever against management. For some software this is case. If you write safety critical software you can be held personally responsible for accidents. In those industries engineer pushback is much more effective.
My background is safety-critical and -- speaking from first-hand experience -- it was truly incredible how much push-back we'd get when raising safety issues.
System tick timer rolled over after 2^32 1ms clock ticks (a little over seven weeks) and the software mishandled it by rebooting and losing control of the process. It wasn't until QA got involved (it turned out they saw it on a long-term test) that we were cleared to actually fix it. It was a two-character change in an inline function!
In some places "cover it up, don't discuss it in email" is the natural response. It really depends on the management and the organisational culture. You can have management who support the engineering team, or you can have management who micromanage and overrule and end up turning a two-day hotfix cycle into ten months of "try soldering a resistor to every single pin on the PCB".
