I thought the point of xG was to show how likely, on average, goals would be scored from a certain position. As such, the notion that some players are better than others at finishing, or there's an R in the month or whatever is irrelevant.
It's misused and misunderstood as being how likely a team was or is to score. That's not what it is. It's just a total, or a stat of a particular chance showing on average how many goals have been scored within the dataset used, from what are deemed by the algorithm as similar chances.
It's more useful as a measure of how good teams or players are, or have been in a game at finishing than anything else.