rapoto and MeiXi:
This is indeed a bug - actually it's not a problem with the PlecoDict viewer software at all, but rather with the ABC database file, or more specifically with the still-very-buggy software we use to convert it to Palm format. This is also why searches in the NWP English-to-Chinese dictionary sometimes refuse to locate words correctly.
This should be fixed in the next preview update, which also will include a number of other refinements to the databases (such as using Unicode circled numbers instead of (1)/(2)/(3) etc, superscripts for homonym numbers, and fixing a lot of the ??? or missing Chinese characters).
rapoto, re your suggestion about mixed character/Pinyin input, we've designed an algorithm to allow this kind of searching, and it shouldn't be too too complicated to implement, but I'm not sure if we'll manage to get coded and debugged in time for PlecoDict 1.0; it'll certainly be in there at some point, though.