I don't really understand your point here. I completely disagree to that being the best play. Objectively, making a play that you will knowingly die to, is most of the time a bad play. Unless, you plan on wiping the whole team or securing a win in those conditions then it is genuinely a skill issue.
The point system (when working as intended) rewards players for actions that help towards their team's success. These type of actions include kills, capping, uber drops, ubering, sentry destruction, etc etc. None of these actions distract from the goal of the game, nor is it an "arbitrary frequency", most if not all of them can attribute to a team's victory. This is due to the actions encouraging plays while also rewarding players for making good plays. Specific goal-oriented actions like pushing the payload, defending points, and winning the round further enforce this.
Additionally, most actions also encourage classes to play to their roles, with good examples being teleports, saps, ubers, sentry destruction, uber drops, deflects, and extinguishes exemplifying this. There isn't a surefire way to exploit the system other than airshot stickies as far as I know. However, even then, airshots being broken don't distract or shift focus in a significant way. Not sure what silly strategies your even referring to.
"Objectively, making a play that you will knowingly die to, is most of the time a bad play."
So, in your own terms, sometimes it is not a bad play? doesn't sound very objective
I also strongly disagree that plays which get you killed are "objectively" bad, most of the time it is advantageous to stay alive obviously but if you make a play that kills a high value player who is significantly more skilled than you where you die afterwards, or a play where you kill the VIP for example and then immediately die, or a play where you cause a med to drop uber and then immediately die, these are all examples of ways you can choose to trade your life for something constructive, and often times these trades do not have feasible ways of surviving.
"None of these actions distract from the goal of the game"
The goal of the game is to win, there are hypothetical situations where focusing on something that gains you points is no synchronous with focusing on something that gets you closer to winning the fastest/most efficiently. One example is to let an engineer build buildings before you detonate your sticky bombs because you know you will get more points. Another example is letting enemies capture control points that you know you will be able to recapture for the bonus instead of defending them.
"most actions also encourage classes to play to their roles"
Trying to win the game also encourages classes to play their roles, typically.
"There isn't a surefire way to exploit the system other than airshot stickies as far as I know."
You get points for hitting enemies with the brick as scout for "airshots" (possible glitch?), you could play a payload map, let your team push up until spawns shift, and then have a friend on the enemy team sit outside of your previous spawn with you as you toss bricks at them, running back inside to get more to throw at them. This could be cycled for the whole round if the player being bricked was a heavy with the sandvich or of a health kit spawned nearby, or even if they were a medic and used the passive heal over time. Could even build an uber for getting past the enemy team >.>
Are these strats stupid? yes, but they are points-oriented, they will get you a lot of points and minimize the danger of you losing points since you are unlikely to be killed so far behind the action. PL_Frontier would be perfect for this by the way because the final point is very defensible and the map is large with spawns that go very far back. Your heavy friend and you could just rack up dumb brick 'airshot' points.
"Not sure what silly strategies your even referring to."
I wrote about a couple of them in a reply to Mr. Pie you can check out if you'd like, I don't believe the list in exhaustive.