Generally speaking, HPC it is a good way to go. The technology gets you to the next level of scale and many of companies that are specializing in and offering these solutions no longer have to worry about being a 'going concern'. These companies have money, real customers and good runway to continue their growth. The key is really the fact that actual companies are using the tech as a critical part of its infrastructure. While there will be another evolution in the technology, it's unlikely we'll see massive changes within 5 years so if CIOs adopt hyperconverged infrastructure in the next year, they will likely still be on point for the next 3 years or so.
It's a the natural next step in the commoditisation of hardware as a response to the various cloud compute services out there. There is I believe still a strong demand for on-premise infrastructure for organisations in many sectors, and specifically where compliance, availability, or workload requirements dictate shying away from public or semi-public cloud platforms. The real value-add with hyper-converged as I see it is the ability to integrate systems and provide a much more easily manageable platform in a more accessible way than has previously been possible (think Vblock or FlexPod where scale is quite staggered). The accessibility also extends to features and availability offerings that typically would require specialists which can now be tackled more easily by generalists which in smaller organisations is the bulk of the technical skill.