The D2 posting system has been "fixed": it automatically replaces permitted characters with the corresponding HTML entities. It strips out any other characters and non-allowed HTML entities. Hence, "fixed"... it doesn't really work, it just works some of the time.
I.e. to enter £...
Alt-156 (£) works in D2 only
£ works in either D1 or D2
£ is stripped out in both D1 and D2; Slashdot doesn't recognize it and strips it out of your post.
That's probably enough of an explanation, but if you care to know the why and how...
The D1 posting system parses your post as 8-bit text. It is not actually 8-bit text; it is actually UTF-8 encoded. Since UTF-8 encodes characters with code points U+0000-U+007F in a single byte, it is backward-compatible for this range of characters; characters above U+007F require multiple bytes to encode in UTF-8, which is why Slashdot ends up garbling them. The D1 system doesn't do any conversion from UTF-8 to 8-bit.
Try it: paste £ into Notepad and save as UTF-8, then open the file a hex editor. The file will be 5 bytes: the byte-order mark (a zero-width non-breaking space, code point U+FEFF) encoded in UTF-8 (EF BB BF*), followed by the pound character (163, U+00A3) encoded in UTF-8 (C2 A3** - which, as 8-bit text, is the characters £ - which is what you ended up with in your post; it appears that you used the D1 system to post the comment).
Note that the £ character is actually code point 163, not 156. Typing Alt-156 produces the pound symbol as a throwback to the DOS code page 437, which contained the £ character at position 156. In Unicode, the £ symbol is code point 163 and can be typed Alt-0163.
* 0xFEFF, 11111110 11111111, mapped into the 24-bit mask 1110xxxx 10xxxxxx 10xxxxxx = 11101111 10111011 10111111 (EF BB BF)
** 0xA3, 10100011, mapped into 110xxxxx 10xxxxxx = 11000010 10100011 (C2 A3)