Good points. I always look at the sentimental value.
When I'm choosing my coins, I don't even check how much they're worth. I do the research on team and their solutions.
Great example of good coin like this is Gridcoin - it was built on top of very famous researching software, became a token of appreciation from scientific organizations to volounteers. Is backed up by 7 years of developement and big partners (such as Seti and other prominent researchers).

However there are traps. Another 'cloud-super-computer' coin, comming out soon is called Golem. Generally it seams like a great idea, block-chain based super-computer possible to be rented and used by anyone with credit.
However is there a market for such solutions? To make it short: No.

Really large computing power is already covered for science by Gridcoin. For private pruposes the code can't be distributed by block-chain, mainly due to the confidentiality of operations. It's also cheaper and more secure to keep a dedicated machine in an enterprise business. So golem, even thought it looks all flashy and shinny - by me is not such a great technology.