The observations, temperature measurements, are evidence. Ideally, any “adjustments” would be reversible, or like I said, benign, like a units conversion. They might also have done something like account for known problems with certain sensors, removing known biases, for instance.
A great number of valid scientific observations are only available second hand. For instance, Ptolemy recorded Hipparchus data, although how Hipparchus arrived at his results is not known and our knowledge of Ptolemy is indirect, from Arab sources, although in the original Greek, which Arab scholars were careful to preserve, though not without accumulating scribal errors over the centuries.
Of course, if Arab scholars could preserve the original Greek of Ptolemy for millenniua, in copied and restored hand written manuscripts, it seems odd that East Angelica University couldn't preserve the world's temperature record for even one generation.
and that can reliably predict future observations. However, they seem to fail even at that.
A prediction of the average global temperature is best done by taking the daily energy of the Sun. Not by taking local weather conditions. The local weather is influenced by Clouds, and seasons due to polar tilt.
The AGT could go up, and yet the 'local weather' could get both hotter and colder than ever before.
Climate change occurs from changes in the Jetstream and changes in ocean currents.
Neither of those items is in any way controllable by man.