Your point about there not being a difference between 5000 and 3000 is simply wrong. It's different a factor of 5/3. How many billionaires would turn down the opportunity to have 5/3 more money? None.
the opportunity to get it isn't costless. most people would be irrational to take a substantial risk of bankruptcy for a chance to have $5 billion instead of $3 billion, given, among other things, the diminishing marginal utility of wealth.
And my conclusions have done nothing but sharpen on the incredible outcome: bitcoin is likely to actually be adopted as a currency by a huge fraction of the world.
for one thing, three weeks isn't enough time to make lifelong investment decisions. if you feel rushed, that itself is probably the result of counterproductive speculative pressures. recall that 'get in before it's too late' is an almost certain hallmark of a scam or at least hypothetical speculation. in non-speculative investments, there's a presumption that the price impounds at least
some rational calculus by the market, so that waiting a month ought not to make a big difference.
as to the particulars, your interesting article runs into a few potential problems that you might want to consider being more cautious about.
first, exponential growth is almost always unsustainable and, in practice, is almost never sustained. relatedly, past performance of any commodity does not predict future performance. to assume that it's as easy (and rapid) to grow from 1000 to 10000 users as it is to grow from 100,000 to 1,000,000 would have been an incorrect assumption for almost any technology, and it's not justified here.
second, you downplay various attack vectors. it's very hard to predict, but many have offered at least credible analyses that suggest (for example) that governments could readily shut down bitcoin. there are also potential internal problems, like scalability and the evolution of the fee structure. at the very least, there are more unknowns than knowns here.
third, i don't see a justification for your claim that there can be only one decentralised online currency. you tie this claim only to the notion of the security of the block chain based on hashing power, but
(1) alternative currencies could rely on alternative proof of work,
(2) alternative currencies could rely on alternatives to the
need for proof of work, and this becomes sharply easier as soon as you relax the design criterion of 'total decentralisation' toward 'decentralised for all practical purposes', and
(3) 'most secure' is never important in the real world; what's important is whether a technology is secure enough for the purposes for which it's being used, in view of its threat models.
(as a rough analogy, the rijndael cypher was adopted as the advanced encryption standard, aes, even though almost everyone believed it to be less secure than serpent and twofish. but it was faster and good enough for all conceivable practical purposes, given the standard's timeframe, threat model, and use cases.)
i'm not sure why i keep engaging these discussions, only to be misinterpreted as a bitcoin detractor. i'm actually trying to help people make better investment decisions, and though you don't have any particular reason to believe me, i have quite a lot of experience in this area. as a matter of full disclosure, at today's rates about 3% of my wealth is currently in a large pile of bitcoins that i mined cheaply early on, though that was not a conscious investment decision and amounts to little more than a lottery win.