Eye witness testimony is unreliable.
Air force intelligence services need accurate aggregate losses of enemy aircraft in a given operational area; they don't need accurate scores for each pilot. When pilots exaggerate, they force intelligence analysts to make statistical corrections. Those require accurate counts someplace, and the only way to do that is have somebody count wrecks and captured and killed enemy pilots and compare that number to pilots' claims.
Your first two points are of course correct. Trying to correlate the sort of data that the intelligence officers needed from eye witness testimony is a task loaded with potential causes of errors.
The problem with counting wrecks and captured enemy pilots is not always possible and the statistical corrections themselves may compound the errors. This was certainly true of the RAF during the BoB when the huge discrepancy between claims and aircraft downed over Britain was closed by the assumption that many Luftwaffe aircraft had come down in the sea. In fact they had not.
The RAF actively discouraged the sort of treatment given to someone like Marseille in Africa and did promote the idea that a squadron was a team, not about one individual. When some RAF fighter pilots perceived (or thought they did) that an individual's quest to up his score was having an adverse effect on overall performance or, worse, putting them at risk, they would say so. This was the case with some who served with Bader for example. I've never heard of such a thing in the Luftwaffe.
To some extent a factor as mundane as the prevailing weather conditions can have an effect. Luftwaffe claims in North Africa, despite some well known fraud, were actually remarkably accurate. Aircraft did not disappear into cloud and crash sites were clearly visible for many miles. This is not the case over NW Europe or the British Isles. Whether an aircraft had indeed gone into the English Channel or North Sea was much more difficult to ascertain with any certainty.
Cheers
Steve