I haven't used the library yet, so I can't say how the code was written. But, as I was studying it with the intent to use it, this note seemed very confusing. Since I2C is a serial protocol, it has no inherent byte alignment issues. It's simply needs a 7 bit address. So, the question I'm trying to answer is, how are these 7 bits aligned in the value passed to a function that takes an address (and, BTW, I assume that an address is a char, or byte value? I'm nit sure, as the docs don't say.)
If the 7 bits are aligned to the LSB end of a char value, then you'd want to leave the MSB clear and you would end up with a value between 0-127 (inclusive.) However, if the 7 bits are aligned to the MSB end and you are to leave the LSB clear, then you can't wind up with a value between 0-127. You'll, instead, get all the even values between 0 and 256 (inclusive.) So, no matter how you look at it, the description is confusing, IMO.