Those who've worked in aerospace manufacturing over the past few decades will realize the truths in the HBR article.
When management perceives their company to be operating in a high-cost arena (very much including all those expensive senior engineers) it's very comforting to imagine the problem can be solved by outsourcing the whole package to a world-class supplier in a lower-cost environment.
The HBR reference to module outsourcing doesn't go far enough. A corollary to the yet-evolving-interfaces problem is that an interface spec robust enough to be "thrown over the wall" to a supplier must be grindingly detailed - which are metaphorically full of devils, of course. Those specs are often written by mid-level engineers (or recent grads) willing to undertake that grind, and once issued, making corrections to them for whatever reason involves project and contracts people who'd rather you didn't.
Perhaps with that in mind many of the major players have formalized their engineering processes with a set of SOPs that when followed will produce a perfect result every time. They include a set of metrics that measure compliance with those SOPs. Thus when an external auditor agrees that the company's processes are being followed everyone can relax, assured that the airplane will come with a safe and happy electrical-system module.
Surprise.