Very interesting numbers. Do you know what the methodology was behind them?
Yes. I calculated it.
Each state gets a fraction of the total number of electoral votes nationally. Divide that number of Electoral Votes for the state by the population of the state to get a voter's value under the EC system in that state.
Under the Direct Election (DE) system, everyone gets a voter value of 1 / (US population).
Divide the former by the latter to get the relative vote value per state, and express as a percentage.
Yes, not everyone votes (age, apathy, registered), but I presume each state's voter / nonvoter ratio is about the same.
Yes, populations change with various effects. I used 2006 population numbers, and assume that any changes average out to generally negligible.
The result is a ballpark view of the relative value of a vote for POTUS in each state. Differences of a very few percentage points are presumed equivalent, at the statistical noise level. It's the >10% differences that are interesting to the discussion.
You methodology is too superficial. Scientific American had some political scientist do an analysis some years ago.
Basically, they concluded, as you do, that the votes of inpopulous states were overweighted. Counter intuitively, so are the votes in large states. Since a small number of swing voters can decide an election in a large state like Ohio, they leverage more votes. A good gauge of the value of a vote is the dollars per voter spent in various states on political advertising.
It's not only how many electoral votes per voter, but more important, the marginal electoral votes per voter. Living in Massachusetts, our votes in the Presidential election are almost irrelevant. No Democrat can win if he doesn't carry Massachusetts easily. (Reagan carried it in 1984.)
Democrat or Republican, if you live in the PRM, your vote for President doesn't count.