2.7V is only partly relevant. Raspberry will detect anything above 1.6V as "High", so having 2.7V (with a closed switch) is no explanation for the malfunction.
Still, you need to understand where that voltage difference of 0.6V come from: either you're measuring the wrong signal, or the actual schematic is different from what you've drawn. In both cases, finding the root cause will likely also explain why the encoder doesn't work.
I see no values for C1/C2: if those are high enough, then an intermittent 0-3.3V signal can easily become fluctuating around an arbitrary value, depending on the rotation speed and the width of the pulses.
Probing the signals with a scope would be the best: multi-meters are only useful for measuring voltages which you know for a fact are constant.