I also want to thank cruachan. Without his hint it might be possible, that I would have spent some more days for this challenge without any result.
Causal for the problems with this challenge is this character, which looks very similar to the ASCII character 'minus' (= 2Dh) but has the value 96h and is not part of the special character set used by CrypTool. This character 96h is part of the original wikipedia article and the plaintext, which is part of this wikipedia article, also contains this character 96h. The length of the plaintext with this 96h character is 1120. While encrypting this plaintext, CrypTool ignores this character (because it is not part of the special characters used by CrypTool) and does encrypt only 1119 characters, as it was apparently intended. But this character 96h has not been completely ignored. It is true, that it has not been encrypted but instead it has been copied unchanged from the plaintext to the ciphertext and therefore the ciphertext has also the length of 1120 (instead of 1119) and contains this illegal character 96h.
If I decipher this ciphertext (with 96h) with CrypTool and with the correct matrix of the appropriate size, all looks fine. Again this character will be ignored, deciphering 1119 characters. Instead this character will be copied unchanged from the ciphertext to the plaintext and I get the original plaintext with a length of 1120.
But wether I use CrypTool or some other software is not the true problem here. From the visual impression I cannot tell the two characters 2Dh and 96h apart and for me, naturally, it was a 2Dh, because 2Dh belongs to the special character set of CrypTool and it is the only character there, which looks like a 'minus'. There was no reason to assume the presence of any 'illegal' characters (not belonging to the special characters used by CrypTool) in the ciphertext. Therefore I assumed a length of the ciphertext of 1120. Now hint number two in the challenge states:
"The length of the given ciphertext fits to the size of the
sought-after key matrix. The plaintext was encrypted without
padding."
Following this hint I tried several different sizes for the matrix according to the length of 1120, but the size of the matrix used by the author is not possible at all with this length. Therefore I could have tried countless matrices of different sizes (according to the length of 1120) until the end of all times. The Queen would say: "I am not amused!"