The effects you are concerned about here do not apply until the system reaches "peak value".
Or a plateau.
if I have $100 of Steem Power, then I must contribute $10 worth of value per year.
Exactly, but this also means that if I can only contribute $10 worth of value per year, I should keep $100 of Steem Power at most and sell the rest. Won't this create downward pressure?
You can hold your value in Steem Dollars. If you expect no growth in the platform then selling makes sense. But those who are active have financial incentive to grow the value, not just maintain status quo. After all, if the combined efforts of everyone involved only enables them to "break even" then there would be a large expenditure of effort and no profit. People work for the sake of profit. As a whole the community will have to produce a profit or die.
Your hypothetical plateau or steady state will never be reached. Steem will change hands until the new owners / voters / users / readers find ways to produce more value than they consume.
Steem is mostly driven by speculative value. This means that any impacts from dilution are insignificant next to changing speculative perceptions in the market. To put all of this in perspective, imagine a company that decided to stop paying for employees, inventory, and advertising. It cuts its expenses to the bone, but will slowly starve to death.
If you are worried about 10% per year issuance then you are short sighted. All living organisms consume energy and must find a way to replenish their energy. Two approaches are to reduce consumption or increase production. Reduction in consumption can only get you so far before you are dead.
So you expect Steem to generate gains mostly through market cap growth. Fair enough. But is there anything that ties market cap and value creation on the platform, aside from speculation?
Would you compare investing into STEEM to investing into start-up shares, with no dividends but a high potential for growth?
That's a fair assessment. They're just like startup shares. Most early stage startups dilute to bring more resources to a company. The goal is for the dilution to bring more value overall than the amount that is diluted. Older investors get a smaller % of a bigger pie, but the value of their share increases. If you look at Steve Jobs or any early founder they have significant stake early on, but over time they have a much smaller percentage. Jobs had less than 1% when he passed away. He did sell a significant share before leading Apple to a comeback so his stake is a bit skewed. Nevertheless the Jobs example does illustrate the nature of early stakeholder dilution. When you're in a garage with a few co-founders you start with 100% shared amongst the team. Overtime you dilute and if you get really big that percentage is probably going to be far less. Early on for STEEM it's all about the network effect just like it was for Facebook & Twitter. Revenue can validate the value a platform, but it's not necessary especially early on. Advertisers will eventually want to buy STEEM to advertise and what they will pay will be an indirect representation of revenue.
@officialbitcash: sorry, it seems we reached maximal comment depth, so I can't reply to your comment directly.
I am not so convinced that advertisers will be interested in Steem. Obvious ad posts will likely be downvoted (by regular users or by competitors).
Regarding the startup analogy, I realized there is something that startups have and Steem does not, though: the eventual promise of (real) dividends, in the long term. I suspect this is a part of what drives the valuations of startups up, even though they often do not pay dividends in early stages.