Serious problem with not recognizing transpositions

As per title. I'm not even talking about moves that take longer to mate. These forcefully transpose to the solution in the same number of moves.

Case #1:
en.lichess.org/training/37823
Solution starts 1... Qe3+ 2. Kh1 Qe1+
The alternative 1... Qf2+ is falsely marked incorrect even though 2. Kh1 Qe1+ forcefully transposes.

Case #2:
Solution starts 1... Rxh3+ 2. gxh3 Qg3+ 3. Kh1
The alternative 1... Qg3+ 2. Kh1 Rxh3+ 3. gxh3 again forcefully transposes in the same number of moves.

whoops forgot the link for Case #2: http://en.lichess.org/training/13542

http://en.lichess.org/training/18369

Another one, Ke4 and Ke3 do the same thing but only Ke4 is counted.
Too many of these garbage problems. Wouldn't be that difficult to improve the puzzle generating algorithm to prevent this.

Hi Thibault.
I don't understand what that link is supposed to address. The problems are obviously faulty (I even played them against stockfish) and I posted about it in the right place haven't I?