What you appear to be describing is that your base game has a 16 bit font (mainly as 8 bits is not enough for any real amount of Kanji) and you are having use 00 between each letter, this sounds very much like U16 unicode so if you are tapping in English or something else then consider changing your encoding to be U16 unicode.
The problems this causes are twofold
1) Storage within the game
2) Memory use during runtime.
I would argue 1) is not such a problem with modern consoles as you basically have unlimited space when it comes to text type things. It is also where you would use compression.
2) is still a problem on modern consoles though not half the one it was back on older consoles.
There is a third problem which is kind of a combination of both (the pointers/text format can not handle the resulting file size) but that is not such a problem either.
That is somewhat beside the point though as either way (assuming you actually need to do it) you will have to do an what is usually termed a 16 bit to 8 bit encoding conversion. Traditionally (and may still happen here) this involved lots of assembly hacking to get the game to read and decode the text as 8 bit entries and was probably the hardest hack in text hacking with only variable width/line handling font hacks and text box changing coming close to it. More recently I have seen games have extensive font/text encoding handling engines though these also bring further problems in some cases as you get to hunt down and change everything that needs changing,sometimes it might be a setting somewhere and if it would be on any console then the PSP would be high up on my list of safer bets for it but I would not go in expecting such a thing.
The short version is yes it is possible, I just discussed the two main methods of seeing it happen and a lot of the time it is one of the hardest hacks in text hacking.