We do not need to recreate the wheel. As we all have seen there are some standard signs for chess analysis. These signs are simple and standard and every normal man understand it. The language must be like that. (Brilliant !!, Good !, Interesting !?, Dubious ?!, Mistake ?, Blunder ??) Any person or programmer who has decided to make Chess.com analytics had to use that but they preferred to use some dramatic phrases like BEST MOVE. You can see such problems still exist in the puzzles even after improvement of calculating the scores, because you must find what ever move the players have done which are not the unique. Sometimes you can change the first move with the second one. Two reasons I guessed. One is to use real played games as puzzle but puzzle must have one unique answer for the first move then it can be a few branches and all the branches include score. Second is that the computer has found the answer must calculate all the right answers then makes score based on that.
Therefore I suggest chess.com programmers to work with chess standards.
I've been saying for ages that they need better programming.
In my understanding, an excellent move must be better than best move but the frase "Best move" must be by definition the best move, even better than excellent. Which is it????
I have seen it post really bad moves as "excellent", whereas the analysis tool's "best move" is rarely a losing move, even if it isn't the best move. So ignore the adjective "excellent". It's meaningless because it doesn't even always mean "OK".