There are ongoing debates about how to make projects that raise funds via ICOs more accountable. One of the prominent threads is that by using smart contracts, a trusted group of people can control the release of funds to a project, thus keeping the developers honest.
In reality, the problems encountered in the ICO market are not new. When greedy but naïve investors enter a poorly regulated or unregulated market, bad things usually happen. Remember the Vancouver Stock Exchange? Pump and dump schemes, mining (!) company scams, directors absconding with money (sometimes tens of millions of $) were regular occurrences.
As to the gradual release of funds: one can already open a bank account for a project, have 2n-1 signatories and set things up that any transfers require n signatures. Smart contracts may make this simpler and more transparent, but will not change the fact that the signatories will have to decide somehow based on real-world facts. Banks have traditionally used a similar approach in project financing, where funds are released in “tranches” according to a progress schedule. Tracking such progress is relatively straightforward when it is an industrial or public works project: when building a bridge, it is easy to verify that a new section has indeed been built, whereupon a new credit tranche is released.
Unfortunately, evaluating progress in a software project is not that easy. The development process of complex software is almost always chaotic. There is trial and error, a direction is taken and then discarded, a new module is written, but has to be rewritten later. In addition, in small teams internal disagreements can have massive impact on progress — but that is of course something that applies to any start-up.
Let’s not forget that projects raising funds via ICOs are start-ups, and most start-ups fail. This will not be different for ICOs. In fact, some blockchain projects seem to have a particularly high probability of failure, as the focus is mainly on technical aspects to the detriment of solid management and marketing talent. There is also the the belief that prominence in the crypto-space somehow makes a project more likely to succeed: the ability to raise $millions is being confused with the ability to execute a commercially viable project.
Bottom line: the problem to be solved is not, and will never be, technical.
Using smart contracts may help, but mostly the ICO market needs to start self-regulating, which is already happening with the emergence of ICO evaluation services. With standard due diligence, the most egregious cases of fraud will be easily eliminated, but no smart contract will ensure the success of a project.
Coming back to the idea that a trusted group of people can control the release of funds to a project: one possibility would be to encourage the creation of independent audit groups. A project doing an ICO could commit to retaining the services of such an audit group, which would then be given access to the project’s github and provide an ongoing evaluation of progress. These evaluations would be public, thus making projects more accountable. Whether it would also make such projects more likely to succeed is another matter…