I think the place where big companies go wrong is by having teams that are too specialized. If you're the Signal Integrity team, the only thing you care about is signal integrity. You should just say no to anything but the most conservative arrangement possible; you're responsible for Signal Integrity, not profit or time to market! Block the project until the signal integrity is as good as it can be, that's your one and only job.
The even bigger problem is that the existence of the signal integrity team means that signal integrity experts aren't embedded on the product team. The product team can happily say "they're handling that", and cut whatever corners are necessary to get their $25,000 bonus, not caring at all about the tickets the Customer Service Team will have to handle as a result of poor signal integrity, the weird bugs the Software Engineering Team will have to work around, and the terrible reviews caused by a computer that randomly flips bits in memory.
Engineering is a discipline about balance, and I think the balance has to be done inside the team. You need to make a conscious decision to trade off a lower BOM cost for more customer service tickets, and the team needs to own that decision and plan for the future under the constraints of that decision. (Instead of sending out an order for capacitors, send out a req for more customer service.) If there is someone in the "out group" that you can pass your concerns off to (i.e. that makes the cost of your balance decision cost 0 for you), then you'll never make the right balance for the product and company as a whole. You'll just maximize for your own gain. From there, you ship your org chart, the company becomes entirely about politics, and you're out of business... just like DEC apparently.
(I guess I also want to comment on the article as a whole -- being bitter about too small a bonus. Something similar happened to me and I sure was bitter about it. I actually quit over it. It all sounds very rational to me, knowing how humans work, and I can't believe management would do something so stupid. It's the kind of thing that kills your company, and yup! it killed them! If there is one constant in the field of engineering, it's that we forget the mistakes of the past much too quickly. Except Therac-25, everyone remembers that one, but integer overflows are still killing people to this day. Sigh!)