FORTUNE — There is little doubt that today’s smartphones offer room for improvement. Ubiquitous and beloved though they may be, the portable devices are fragile, offer limited choice, and generate a growing mass of e-waste that’s been called nothing short of a “global time bomb.”

Google’s Project Ara aims to change all that. With a modular, open-source approach to smartphone design, Project Ara phones will be based on a structural frame — dubbed the “endoskeleton,” or “Endo” — on which to attach smartphone modules of the owner’s choice, such as a display, keyboard, camera, or even specialized components such as a glucose monitor. In other words, you design the phone you want, and don’t pay for parts that you don’t care about.

The concept is notable for its customizability and overall do-it-yourself appeal, but modular phones are also anticipated with much enthusiasm for the reductions in replacement costs and e-waste they could bring. Fry a component? No need to exchange or throw out the whole phone — just replace the part that’s broken. The result, as some have suggested, could be that the modular smartphone is “the only phone you’ll ever need.”

That’s the thinking, anyway. The reality, however, is that modular phones have a long way to go. The first Project Ara phones are expected in January for around $50, but a lot has to happen first. And even if the pieces do fall into place — so to speak — it’s not at all clear that consumers are ready.

‘People are sick of throwing out their phones’

Google’s modular approach may be the best-publicized so far, but it’s not unique. The Chinese manufacturer ZTE offers the quasi-modular Eco-Mobius. And preceding the October announcement of Project Ara was the launch of the Phonebloks project, led by Dutch designer David Hakkens. Created to address the e-waste problem, Phonebloks envisions an “app store” of sorts for hardware through which users can buy new and used components as well as sell their old ones.

Phonebloks has since thrown its support behind Project Ara. Before that, it conducted an online campaign on “crowdspeaking”platform Thunderclap, where it garnered more than 950,000 supporters and attained a social reach of more than 380 million people.

“We didn’t know what the demand for this was — that’s why we did an online campaign,” Hakkens told Fortune. “Turned out quite a lot of people are pretty sick of throwing out their phones every couple of years.”

‘Manufacturers aren’t interested’

There has been no shortage of critics for the modular concept. (A taste: the writer John Brownlee, who penned a piece for Fast Company shortly after Phonebloks’ launch entitled, “Why Lego Design Principles Don’t Work On Smartphones.”)

Hakkens admits there are significant market challenges to such a phone.

For one, “current phone manufacturers aren’t really interested in this phone,” he said. “Currently, they are in charge and decide which components are in the phone. They prefer to see an open market where they can sell straight to the consumer.”

With regard to Ara, there remains a lack of clarity around the project since Google sold much of Motorola to Lenovo in January 2014. “Who’s going to do it?”asked Ramon Llamas, a research manager for mobile phones at IDC. “Is it a Motorola thing? A Google thing? This has Google’s fingerprints all over it, but with Motorola heading to Lenovo, there’s reason to question. Does Google want to get into the hardware business again?”

Even if it does — which is unlikely — it will need an ecosystem of partners to make modules for the phone, said Peter Semmelhack, founder and CEO of Bug Labs, which makes a “Lego-like” hardware platform of its own for digital devices. “Every module is a product with its own life cycle and road map,” he said. “It becomes a geometric problem.”


‘The countervailing problem will be cost’

As for retailers and carriers, a modular device could — albeit briefly — offer a modicum of much-needed differentiation, Llamas said. But the investment required could be significant.

“Think about training your sales force, how much inventory you’ll have to carry, all the different colors and designs you may have to have,” he said. “The last thing you want is to hang onto inventory that’s just not moving.”

And then there’s the question of whether such a device is feasible from an engineering perspective.

“I’m not an engineer, but I know if you take two objects and slide them on top of each other again and again, eventually they won’t lock as well as they used to,” Llamas said. “For this to really come together, the parts have to be able to withstand shock, drops, the elements, and that’s a huge thing for me.”

One of the most compelling arguments? Cost. The inability for consumers to service their own devices today is hardly an oversight; it takes a variable out of the equation.

“I have a soft spot in my heart for what they’re trying to do,” Semmelhack said. “They’re trying in essence to provide consumers with what they’re missing right now, which is choice.” At the same time, “the countervailing problem will be cost. Modularity comes with a lot of complexity on the hardware level, the software level, the mechanical design level — even aesthetics. In a mass market, cost is probably the most important factor, and the only way to get the price down is to make the same thing in large quantities.”

With the modular approach, Semmelhack said, “you lose that scale.”

‘Moto tried this before’

The modular concept isn’t without precedent, most notably in the PC category.

“We’re seeing, through Ara, an attempt to make smartphones much more like PCs in the sense that when you go to Dell or HP or Lenovo, you get to tell them what to put in -– processor, memory, disk, display, etc.,” said Jack Gold, principal of research firm J. Gold Associates. “Ara is an attempt to take this model down to smartphones and probably tablets eventually.”

Will consumers bite? Llamas said it’s too early to tell.

“It’s hard to build demand when there’s no product currently out there,” he said.

Gold said appeal could be generational. “Being able to offer a ‘customized’ phone, particularly to twenty- and thirtysomethings, could be enough of a differentiator to make it very popular,” he said. “Of course, Moto tried this before, with marginal success, albeit with far less customization.”

Semmelhack sees potential in certain industries, rather than the broader consumer market.

“There’s all this emphasis in mass-market solutions to try to find ‘the killer app,'” Semmelhack said. “There are some out there, but there’s also a category of lesser killer apps within communities and subcategories — medical is a great example.”

With a modular approach to smartphones, he said, it would be more feasible to address users with diabetes or juvenile asthma, for example, in economical ways. “Say there are 100 markets with idiosyncratic needs,” Semmelhack said. “That isn’t economically feasible to serve in a mass-market way.”

Done right, it can make economic sense for Google or any other phone-maker. “Instead of selling 10 million of one thing, from Google’s perspective, you sell a million of 10 things,” Semmelhack said. “You can do it economically because you have invested in the platform.”