Tomek's method is in accordance with the rules. The only way of preventing such action is multiple testing on single test-cases, and this won't be available in SPOJ for another couple of weeks. Until then, introducing lots of test cases for new DASM problems becomes a necessity.
The only question that remains is: what should be done about the existing problemset for PHRASES (SOPARADE is not really a problem, since Pascal Zimmer's solution seems perfectly ok). Oh well, its my fault, so I guess I will have to think of a way out. I won't increase or change the test data, unless everybody who has solved the problem agree to this...
Doubling the score was used in previous editions of the league. This year, it doean't seem to be such a good idea. I guess the bonus will be removed or decreased starting from the second problemset onwards.
BTW. In the future, we are planning to use a slightly simplified version of the formula from the league to generate the main spoj ranking. The score will most certainly not be doubled there.