Hi ttookk,

well. When you take a look at some other "decentralized computation" programs, there is no real advantage: some allow you to host your website on the network, some let you use virtual machines on other people's computers. All tasks that could be done more efficiently (both in terms of speed and cost) if you just rented an Amazon AWS instance. The beauty of XEL's design is that you are not just using crypto-currencies to purchase resources from someone. The key difference is, that you are putting your tasks on a billboard for anyone to grab, which causes a huge competition between everyone (not one virtual machine guy, but literally anyone) who is running a computation node. So let us say, I host my "NodeJS" service on one of the other projects for example, I get exactly one dude (or a couple, but certainly not all) to do that for me. In XELs case, you potentially start seeing thousands of nodes fighting simultaneously for being the first to solve your task. That is one of the major issues of quite a few projects - they seem to encourage people to work slowly lol ... I mean why would you run a fast VM or provide a good upload bandwith for that site while you also can just run it at very low speeds  (and potentially get paid more)?

This principle is not new, BOINC does exactly the same with the little exception that there, you rely on volunteers with an ideologic agenda. This makes it hard to convince people to work on your simulation for your homework assignment

By the way, have you looked into zk-Snarks? Coda(https://codaprotocol.com/) is trying to use it as a means to verify the blockchain with very little ressources. I haven't really looked into it, but I was wondering whether this could be a way to prove the validity of complex computations without having to reperform them, since that is pretty much what they are attempting.