Yeah I understand all that and in theory it works well but when put to practice it does not always work well. SBW would be a good example with a really high tackle rate, but he doesn't bring much else to the table. Stats can obscure an impact a player has on the game, they might not have as many line breaks for example, but they might make a game breaking decision that resulted in a try in another area that another stronger line breaker would not have made. Conrad Smith is a good example here.
When you put the stats up between Kirindrani and Conrad Smith, Kirindrani looks the far superior centre and player, but we all know that's not the case.
But I can use tackle stats to argue Sonny Bill Williams is a more accurate tackler than another player, over the period of a series.
"Best" is of course a relative term. Because statistics compiled in rugby tend to be pretty limited, you can only try and use the limited picture they provide to demonstrate an aspect of one players game.
In regards to Conrad Smith vs Tevita Kurindrani - I think you would be surprised. Again taking the stats of those two players (based on Super Rugby):
Conrad - Kurindrani
Minutes played: 1,240 - 1,157
Tries: 3 - 3
Carries: 111-104
Metres: 506-400
Clean Breaks: 10-12
Defenders Beaten: 21-29
Offloads: 12-9
Assists: 1-1
Passes: 118-27
Openplay kicks: 8-1
Tackles: 117-104
Tackle Percentage: 88.6 - 79.4
Turnovers Conceded: 8-9
So as you can tell looking at these stats over the space of a season - what would appear instinctive (Kurindrani would dominate in aspects such as ball carrying / metres) is not necessary the case. If I was to dissect these statistics I would argue based on these Conrad Smith is the more accurate defender, and has a better passing game. Based on your perception I would argue he has a very underrated running game.
Of course it doesn't tell you who is the better decision maker, who provides better on-field leadership - and you have to take into account performances are often relative to the makeup and structures of the team. Unfortunately for these things you either have to go with anecdotal evidence or better yet - match footage to demonstrate a point. What I hope this has demonstrated however - is that what appears intuitive isn't necessary fact. Yes, I agree once again that they are a crude measure of a players attribute: but it is better than nothing.