the annualized growth numbers for the larger buckets (everything above 1 btc) are LOWER than the annualized growth numbers for small balance buckets (less than 1 btc), so not sure that your conclusion is accurate

for anyone who really wants the data...look at gbianchi's recent posts here on bitcointalk.  He posts on an italian forum here, but all the data is there, and he seems to update more frequently on that forum (ie the last few days)

on a side note (this might not be the correct forum for it), but i've done a separate regression analysis on active wallets vs price from June 2010 to Jan 2013...with the idea being that after 2013, the willy bot artificially inflated prices.  This actually shows up in the data as well.  There is a solid line of best fit on active wallet vs price graph from June 2010 to Jan 2013.  After Jan 2013 this relationship breaks down.  Now maybe it was metcalf's law taking affect after Jan 2013...or maybe it was the willy bot.   My model assumes no metcalf's law, and only zipf's law (which is a 2x increase in price for every 2x increase in users...as opposed to metcalfs law which is a 4x increase in price for every 2x increase in users).  My models shows a fair value of less than $100 per coin...but that fair value is growing at 50-100% per year.  Anyways...more details here:

http://www.reddit.com/r/Bitcoin/comments/2h8he9/longterm_correlation_btc_users_vs_btc_price/

i've been in many discussions on this exact topic and have yet to here a good response.  Even peter todd's double spending could be prevented with a simple smart wallet that checks for and flags any transactions that don't come from "vanilla" sources.  If there is anything even remotely suspicious, then the transaction is flagged, and the customer is not allowed to leave the store with the product.

Anyone with more technical knowledge care to refute satoshi here?

wouldn't it be fairly easy for a merchant wallet to check for this type of thing.  If any of the above scenarios are true, then the customer would have to wait for full confirmation.  Basically, wait 10 seconds, have the merchant wallet check for double spends.  10 seconds is enough of a head start that the first transaction should be propagated to enough of the network to make double spending unpractical.

Satoshi addressed this here. Are there any flaws in his plan?:

https://bitcointalk.org/index.php?topic=423.msg3819#msg3819

so based on your active addresses growth, the question becomes what will the relationship be going forward.  Should we use zipfs law, metcalfs law, or the fit that you've constructed in the thread?

as i understand it, zipfs law basically has price growth growing in a linear relation to user adoption which is fairly conservative.  Using your active address chart, we get a prices

Date          Number of act adr.         Price/Coin
1/1/2016    10,105,263                   $3,115.79
1/1/2018    42,548,476                    $13,119.11
1/1/2020    179,151,479                $55,238.37
1/1/2022    754,322,020                $232,582.62
1/1/2024    3,176,092,716              $979,295.25
1/1/2026    13,373,021,965             $4,123,348.44