Cloud mining as an industry is a curious one to me - externally, they are some of the most reputable establishments. They have influence in the media, they are seen as early enterprising successes by businesses, and they have played a role in helping securing the Bitcoin network in all these years. But internally, they suffer from a long track record of unhappy customers. Perhaps unfairly, as those unhappy tend to be the loudest and those in profit choose not to speak but in general, they could improve on a lot of things to improve their standing. This forum, in my observation, has many who agree cloud mining is unprofitable. Personally, I think they are a borderline rip off, targeting newcomers unfamiliar with mining.

In any case, you're on the right track if you're thinking to remodel. There are at least two huge projects (Hydrominer and Gigahash) that are trying to do something similar - managed mining (but in my opinion is just a more accurate way of saying cloud mining. They tackle two points you yourself bring up:

1. Selling electricity instead of hashpower. Electricity is more predictable especially if used up in shorter periods, rather than the typical annual or lifetime contracts sold by cloudmining companies. You buy X amount to run a rig for X time - and quickly you see profit or loss.
2. Improving transparency - absolutely! The most misleading thing outside of vague contract terms is the profit calculator. If you could provide transparent operation costs - electricity consumption and price per unit, blocks found, fees earned, Bitcoin liquidated, rental, maintenance costs. Why not? I would then say that if people bought "shares" of the cost, they then share in the profit/loss. No need for fixed formulas of costs earnings that probably aren't true anyway.