That sounds good on paper, but it doesn't account for a few things, first and foremost is server capacity. Adding that just added upwards of 8000*3 calculation per second across the world. Even if you assume perfect distribution, this will cause additional load, resulting in having to lower the overall world capacity at a time. This may be something you are willing to live with, but given the previous problems of 1017 etc, it could be more problematic than that. Also, said calculation still has to account for average latency vs lag spikes, which makes it tricky to get right. The perfect calculation will still have gaps for people to exploit, just due to the poll time.
I agree this is the general approach they are likely to take, but I don't agree that it is easy to implement right without causing other issues.