This is a common mistake (or proposal) I have seen in Esperanto; I have recently heard the word "alies" when listening to ĈRI, and I have also encountered it in Hasegawa Teru's (Verda Majo's) En Ĉinio Batalanta:
Aliel ol en Japanio, ĉi tie mi ne bezonis konsideri la ideon kaj socian situacion de kunparolanto.
There is also a PMEG page about this issue:
Multfoje oni proponis aldoni al la tabelvortoj la antaŭparton ALI- kreante la novan serion aliu°, alio°, alia°, alies, alie°, aliam°, alial°, aliel, aliom°. Iuj eĉ praktikas tion uzante precipe la vortojn aliel kaj alies.
I would say the problem is well-treated in that PMEG article, but I'd appreciate an English-language summary of why this is considered incorrect.
It seems to be such a natural mistake, so natural that many people come to it spontaneously without having seen it before, and others try to defend it and use it purposely. For this reason I think it's an important issue to highlight and explain, because it seems like a possible hotspot of natural evolution and unintended development of irregularity in the casual use of the Esperanto language. Since Esperanto is a living language, I feel it's not in anyone's power anymore to stamp out natural growth. But because Esperanto is often touted as a completely regular language this seems to be a bit of a glaring contradiction to that claim. Of course "ali-" tabelvortoj are not official, but they seem to appear commonly in day to day writing and speech, intentionally or not.
So, why are these ali- tabelvortoj considered to be incorrect? And, if it doesn't convolute this question too much, has anyone else seen examples of irregularities like this in Esperanto, whether embedded within the 'official' Esperanto language or sliding by undetected in the flawed, natural, everyday speech of its speakers?