even if 'You' correctly substitutes for 'I', your would have 'You am..." which is not correct.
if you have 'You are...", whether the substitution produces 'I are...' or 'Me are...", neither are correct.
The program would require a section to identify the use of 'you', and a method for the correct substitution of the 'you' PLUS a modification of the verb that follows, when required. But a section of code like that might end up being lengthy.
Whereas if the data list incorporated more detailed strings like:
s$(x%) "i am / you are"
the substitution would be more straight forward.
But this string structure won't work with the current program because of the additional spacing after 'i' and 'you'. Might be an interesting challenge...