Google thinks its methods for developing and controlling a software ecosystem can work in hardware because it's convinced of its own competence and acumen. Years of consistent success will do that to a company.

I review phones for a living and I can tell you that sort of approach just doesn't work with hardware. Even modular hardware that might be expanded or upgraded down the line must demonstrate value on day one.

Google's key selling point around Ara will be greater versatility. You'll be able to plug in bigger speakers, extra batteries and displays, better cameras, vanity modules, and more niche parts like glucose meters. That's cool. A great deal of Project Ara is cool, but it just doesn't compel anyone to spend extra money on it. People who want more battery life or their own aesthetic signature can just get cases for their phone, while those who want a better camera will just buy it with the phone — because it's such an essential item.

Despite building expandability into the very architecture of Ara, Google is going to integrate the most basic components into the device, including the processing system-on-chip and the display. So no, the eternally upgradeable device will not be coming from Google. Your RAM, processor, and screen are all fixed, just as with most other Android phones. But unlike the others, Google's device will be thicker, to accommodate the extra connectors, and generally less space-efficient because of its physical fragmentation.

Take a look at the Fairphone 2 for a good idea of what Ara might look and feel like. We all like to idly praise the Fairphone for its ethical ambition, but I don't see many of us rushing out to replace our slick iPhones and Galaxy handsets with a thick slab of hard-to-love plastic. The problem with modular phones is that you have to trade away thinness and structural integrity permanently in exchange for only an occasional enhancement in functionality.

Moreover, the whole Ara modularity bit has just been recreated by an iPhone case from Otterbox. Like Google, Otterbox has a bunch of partners like SanDisk and Square making accessories for its accessory. While we all might like the Project Ara smartphone as a cool gadget with fun interchangeable parts, it doesn't truly offer us anything special or unique, and we're unlikely to invest our money into it the way that Google has invested its own resources.

And why should we? Google has shown itself to be impatient, mercilessly killing off projects that fans might favor, but which no longer serve its interests. When Samsung pressured Google to dump Motorola, Google did the expedient thing and kept its biggest Android partner happy. Google Glass, the Nexus Q, Google Reader, Android@Home, Google TV, Gtalk, the list of demised projects is varied and long.

We've seen Asus try and fail with its PadFone series of phones that jack into tablets — each one demanding a separate tablet, no intergenerational compatibility — and LG's new G5 modular phone comes with no concrete promises of forward support. From today's vantage point, Google is basically asking us to spend money on extras for a phone that won't be as pretty or as well designed as its more integrated competition.

Google's Ara project is not defined by a single piece of hardware, of course. The ultimate goal is for an entire ecosystem — organized by Google as the custodian but populated by big-name module makers like Panasonic, Samsung, and Toshiba — and a plethoric variety of options. The problem? When (or if) Google releases the Project Ara phone next year, it will be the only device that those modules can plug into. For me, Ara's best chance to demonstrate value lies in the versatility of the modules, not the phone itself. If I can switch my Ara camera between a compatible phone and tablet, or if I can take my additive Ara speakers off the phone and attach them to my IoT-connected wall mount, then I might start caring.

Google's big misstep with its Google Glass smart glasses was to just get the hardware out into the world and hope that compelling usage scenarios would emerge. Again, that might work with software, but with costly hardware it's just a recipe for limited adoption. Google Glass was also constrained by the brevity of its battery life, and so it's now on hiatus until Google's new hardware division decides that it's truly ready for the world.

In its present state, the Google Ara smartphone just doesn't solve any real-world consumer problems. It's fundamentally a device about extras, but we've not seen what most of those extras can do in order to judge it fairly. And because those additive modules won't have any other device to slot into, the challenge for Google is to essentially prove that it can give us something more than a smartphone. At least to get non-geeky consumers interested at the start.

I think that's ultimately the issue I have with Google's big Ara announcement. What we have today is a phone packaged up in the feel-good vibes of a long-established modular gadget fantasy and the promise that all the extra bits for it will be totally radical. Maybe it's the cynic in me, but whenever a product's biggest promise is also its biggest unknown, I find myself deeply skeptical.

Verge Reviews: LG's modular phone, the G5