What Golem/iEX/SONM do better than Gridcoin is focusing on establishing a direct 'crypto:computation' market. Anybody can potentially come forwards with token and direct computation power to a compatible task they desire completed where as with Gridcoin rewards are distributed to the volunteers of a BOINC project without the project administrators having to pay.

A 'GRC:Computation' market could potentially be built into a Gridcoin pool (open-source kit WIP), or perhaps directly into the Gridcoin reward mechanism through a proof-of-burn system (directing more rewards to project A over project B proportional to GRC burnt against project burn addresses) or directly within a commercial BOINC project.

It's a reality that GRC has distributed its tokens against verified BOINC computation without a premine/instantmine nor ICO, where as iEX/SONM/Golem all performed ICOs upon the ETH platform. By raising millions through an ICO you have that large initial marketcap & you have a lot of money to hire marketing experts; raising a large amount of money through an ICO isn't a sign of a successful project though (note that none have any deliverable in production yet? It's all speculative).

There's no proven compute capacity behind iEX/SONM/Golem as all asset holders are investors, where as GRC was distributed to individuals with proven computing power.