out of curiosity I worked out the encoding, at least partly. It takes the last 4 bits of the first byte and the last 6 bits of the remaining two from each triplet, so we can directly decode like this:
cdecode[s_String] :=
FromCharacterCode@FromDigits[
Join @@ MapThread[IntegerDigits[FromDigits[#1, 16], 2][[#2 ;;]] &,
{StringTake[#, {{2, 3}, {5, 6}, {8, 9}}], {5, 3, 3}}], 2] & /@
StringTake[
s, {#, # + 8} & /@ Range[1, Floor[StringLength[s], 9] - 8, 9]]
StringJoin@cdecode["=E6=B5=8B=E8=AF=95=E4=B8=80=E4=B8=8B"]
same string
No doubt URLDecode is the more robust way to go. Note there are 8 bits that have been ignored here. Presumably the 'E' signifies the start of a 3-byte code - that's ignored here an should be checked.