Its expansion was limited due to state sales tax laws, and Amazon's historical attempts to avoid customers paying sales taxes. Now that they are largely collecting sales taxes they can expand into more local services like Fresh. Most of these laws changed in the last couple years.
Even if they hadn't expanded beyond Seattle, Seattle is their "home market" where lots of their employees live and a rather good market for a wide-availability grocer. Fred Meyer stores nor its QFC sibling hold a candle to any halfway-decent actual-Kroger from back home. Safeway is just Tom Thumb with a 20% markup. This isn't exactly the Dallas/Fort Worth grocery region, with 11 (yes, eleven) full-service grocery chains. Besides, having innovation on a local scale probably matters as much to Amazon as innovation on a national scale because it gets them shopping data and a "test platform" that can serve as the basis for expansion.
Why do they have to go big all at once? What other grocery delivery services have even tried that in the past 10 years?
Freshdirect is NYC area and Philly, Peapod has a few markets but they are part of a grocery chain so everything but the last mile is take care of logistics wise. It seems like figuring out how to do it at scale and be competitive is exactly what Amazon is doing. If they can make it almost as cheap but have compelling additional features then it helps them roll it out wider.
Instead of "Market Research", and "Building Gadgets" - why not just refer to Amazon's continuing work in Grocery development as "Product Development" - eventually they may hit the right mix of devices, software, services, and pricing that allows them to heavily invest in other markets.
They can get it right (figure out how to increase order sizes, order frequency, margins, etc.) and then expand; or they can expand and try to figure all of those things out.
With their main business they chose to grow as fast as possible and are still figuring out profitability. When you're selling electronics with 50% markup that works. Groceries, which have <5% margins, not so much.
In Santa Monica and Culver City in Los Angeles market I usually see 2-3 amazon fresh trucks on my daily commute to and from work. It's rare anymore that I don't see one. I've also seen several companies that are using them to stock the company fridge with things like milk. It looks like it's taking off fairly well anecdotally. I don't know if there's been any numbers published though. The biggest hurdle I've seen with it is that the site for it is woefully underpolished. It's not buggy but it tends to be slow, and a little obtuse to browse on at times. If you have something specific in mind it's really easy to find it, but if you wanted to say just look at the different types of potato chips they have, it can get a little unwieldy at times.