Here's a little spreadsheet that tests Benford's Law against the Fibonacci Sequence:
I'm sure that the effect is mainly to do with progressions, with the Fibonacci sequence progressing by 61.8% from 8 to 13 to 21 etc. If I tested against a series that progressed by 10% with each number I'm sure we'd see similar results, which stands to reason really as it takes more 10%'s to rise from 1 to 2 than it does from 8 to 9.
If we tested against a series of say FTSE closes since 1984 when it was 1000 or so, then I'm not sure we'd see as pure a result as we would have conflicting forces of progression from GDP/money growth and random walk.
Where this ties in with Greek financials I'm not sure, except that it's a given that the ability of politicians and bankers to lie gets 'progressively' worse as a desired target looms.