There are still plenty of them around and thus the libraries to read and write them are still (and ever will be) supported.
However - the J-headed XML fashion followers decided to "modernize" the plist and thus created a really lame and bloated XML-ization of them. So, while dictionaries used to be written:
{ key = value; key2 = value2; }
Now we have to have something like
\n<dict>\n <key><string>key</string></key><data><string>value</string></data>\n <key><string>key2</string></key><data><string>value2</string></data>\n</dict>\n
See, lots more readable (NOT).
The only advantage seems to be the addition of type tags - string, date, real, int, etc. Frankly I was happy with "everything is a string".
Anyhow, I still use NS style plists all the time. I have parsers in Java, C++, and Smalltalk. You can write a recursive descent plist parser in about a page of code. Once you have one you never write IO code to save data again.