I don't think the early investors understand how ambitious the Golem project is. As a 20+ year IT & programming professional in the mathematical sciences, it seems to me that the technical challenge of making this work is very high. The company understates the magnitude of the technical challenge, and is overly optimistic in its timelines for solving them. There is the very real question: will it work? Maybe, and I hope so, but I have serious doubts about their stated timelines.
Cost/reward ratio
It costs money to run computers (power, space, cooling, maintenance), especially high performance ones that would be needed to be efficient for the kinds of tasks they propose, like CGI. It also costs time to set up, monitor, and manage compute nodes. So the rewards will have to be significant for owners of computational assets to be willing to participate in the network. According to the Golem white paper:
...giving providers a clear way to set prices and decide what fraction of their own idle resources they are willing to rent out.
Owners of computing hardware will set their prices. Underbidders will find out their true costs in due time, and adjust their prices.
Competition
The competition is Amazon, IBM, Google, and Microsoft -- all of whom have been honing their services in the cloud computing market for a decade, and they continue to innovate. They all have multiple revenue streams so can be very cost competitive. Scientific users have large government/university supported data centres at their disposal. Maybe these organizations will choose to be participants on the Golem network. The reward potential for them would need to at least match those from their current business/funding models. The inherent efficiency of blockchain payments might actually make it worthwhile.
Tech Support
Technical support for the network users, and their infinite combinations of hardware & software setups, will be a factor. The quality of support they receive (or not) will affect the number of participants on the network. The people working on computational problems for their livelihood have a low tolerance for delays due to technical problems, and word spreads quickly in their communities. Golem's reputation will determine the rate of adoption.
No mention at all was made of technical support resource allocation in the white paper. It appears that they are banking on ease-of-use, that everything will just work -- you know, like Apple iCloud. Apple surely has zero technical support costs. :P
However, Golem is an open source project, and very few of those have company-backed tech support teams. It didn't stop Linux from becoming a dominate platform. It took 20 years, but that was then, and we live in accelerated times now.
Conclusion
Making the ambitious assumption that the technical aspects work as promised: can distributed individuals and smaller organizations compete on cost/performance with professional computing giants? I have serious doubts. But maybe said giants will actually become Golem service providers. Lots of maybes here. As an investment opportunity, I consider this one a longer-term wait-and-see.