From both an educational and quality standpoint I'd argue that engineers should be focused into fairly small niches for 2-5 year stints, simply bc that is the time necessary to become competent in any given niche and how humans learn. In school we learn theory then practice applying it, an hour of lecture and two hours of homework being the usual standard before moving on to the next topic. To become licensed, engineers work in a given small niche four years (fits that 2-5 eh?), get signoff, take a test, and can work within our limited experience. Nobody worth their salt considers another competent after five mins spent solving one problem nor even after a few months solving quite a few of the same.
JMO but I find the notion that small firms offer "broad" experience rather comically absurd, as I do the thought that large firms "box you in." Most small firms dont have dink for facilities or deep expertise (aka experience) internally. They usually don't have budget for many modern engineering tools, a dedicated research dept, and rarely practice process-driven engineering or have much for quality standards otherwise. The joke that they can design anything but nothing well is often true. Dont mistake broad high-level design for engineering, much less broad engineering experience bc there's a big difference in making standard parts function for low-volume products and engineering an optimized part for 1M+. A bit of design routing a COTS hose is a helluva lot easier than engineering new hose stock and custom fittings. Large companies IME place a lot of value on broad experience and push employees in that direction. They usually have rotational development programs for juniors that place them in a design role, a manufacturing role, and a customer support, business development, or other engineering role for a few months each. To climb the ladder into management you typically need a broad background, having worked successfully in the foundry, in customer service, done an overseas tour, been in a "business" role, etc as well as been in product development and faced the long-term consequences of your decisions. At many small companies its common for managers to be only a few steps from the desk they started in. To each their own opinion, but my usual bit of advice for juniors is to spend 5-10 at a big company to learn proper engineering and gain experience with modern technology then try a small company for work-life balance, cultural, or other personal reasons.
As to an "uninspired, disposable, mediocre, and replacable workforce," I have met quite a few engineers who fancied themselves artists or other creative geniuses. None had ever worked in a dedicated research dept nor realized how process-driven those depts tend to be until I shared my experiences.
