I have only done a handful of problems above #100, so take my comments for what they're worth.

I like the new Decimator awards, to direct attention to the mid-range problems and to fill in the award difficulty gap that others have mentioned.

Perhaps this and other info could be used to generate a difficulty rating for each problem, separate from the Solved By count. Difficulty could be based on:euler wrote:For some time now we've been collecting data for number of correct/incorrect submissions so if that information ever goes live I've got a few ideas relating to that. Similarly I've also been recording the average level of the member when they solved the problem. This too will lend itself to one or two nice awards.

Time to reach x (50? 100?) solutions

Subjective rating by people who have solved it

Correct/incorrect submissions ratio

more...

The difficulty rating would be a new way to sort the problems and to define awards. It would make a clear distinction between the newest problems and the hardest problems.

… which leads me to ask –- if the problems are supposed to get progressively harder, and the latest problems are now quite difficult, will there never be another easy or intermediate problem added (on purpose) to Project Euler? If so, I think that's unfortunate.

With a difficulty rating in place, new problems could be of any difficulty level, and the awards for solving the newest problems would be separate from the awards for solving the hardest problems.

If the new problems were of mixed difficulty, it might keep people involved who are working problems more or less consecutively until they hit the wall and drop out. On the statistics page, just eyeballing it, it looks like many people who get a decent start hit the wall around 38, 67, and 109 problems solved.

Then again, there are plenty of easy and intermediate problems already, and people who are willing to work 300+ problems may be very pleased with the overall mix of difficulty.

Thanks again to the development team!