Temperature starting point estimation is off

The board gives consistently lower temperature readouts (by 3~4°C). I do not have laboratory equipment to test this out but it appears the zero raw value is at about 31.0 instead of 27.5. I believe more testing and update to the library is warranted.

No file or test is named. Start by reproducing the board's temperature reading and comparing the raw-zero behavior described in the issue; establish a reliable reference before changing the library, and consider the work done only when the 3–4°C discrepancy is validated and corrected.

Written by the indexing model from the issue text.