Actually some good points were made here. I will just state my suggestions:
Gold, Silver, and Bronze are nice, and I can see that the difference between 3-2-1,
*may* be deceivingly too big. But does it really affect the final result? It seems it does not.
Using slightly decreased ratios such as 4-3-2 or 5-4-3 or even 1-1-1, still allows
a choice of three puzzles (compared to the poorly distributed method in the recent
past with only one choice). So the results are *not* that different as some may think,
because we have enough voters for it to be normally distributed. If the difference
seems bigger, that has *nothing* to do with the puzzles or anything unfair. It is just
a different way of "looking at the forest".
Just think of it from a logarithmic point of view and you will see what I am talking about.
It kind of reminds me the 2-1-0 system they used at (euro) football a couple of decades
ago. Then, they adopted the 3-1-0 system to encourage teams to win. The same way
the current system encourages people to make really good puzzles instead of average
ones. But then, the team which would win the championship, will be the same(!), regardless
if the 3-1-0 or the 2-1-0 system is used. Yes, there can be some really super rare extreme
cases, where a different team would win, but the chance for this to occur is highly unlikely
(in general), and it is much smaller than an "error margin".
So in the end of the day, there is no "perfect system". But one which distributes
the votes "fairly enough" (like those successfully used in scientific experiments)
can be simple enough for everyone to use (and understand) and have an acceptable
confidence interval. And the one we have here, is already good enough for this cause.
Of course, the more votes, the less the (already super unlikely) chances of a wrong result.