So in order to keep scoring winning players in a sensible manner, your scheme would have to work something like:
(1) If you win, you get your winning score, including early victory bonus just as currently.
(2) If you lose, you get the max base score you achieved (with the small extra scaling to final score that the game gives to losing games)
When I suppored this idea, I assumed that win score is the same, and loss/retire score would be the usual loss score, or the highest ever base score (whichever is higher). Not sure if highest ever base score would need to be modified to maintain the incentive to continue. And I expected that the highest ever base score would never be more than an ultimately winning Firaxis score.
The trouble is that for someone who wins the game very late , (2) could theoretically be bigger than (1). In effect they'd get a smaller score than they would've done by just retiring, or doing something to make themselves lose.
I find that to be insane! (I'm not saying it is not true, as I don't know, but if true, it is crazy

). Does the scoring system really do that, or only if you are winning big, then blow it big, and then pull out a win anyway? And how often does that happen?
A smaller problem is that it could encourage players who suspect they will lose to play 'badly' in the sense of temporarily getting an absolutely massive empire that they know they are in no position to defend and will lose very quickly, maybe not even bothering to defend it, just so they get a high 'highest base score achieved'.
That did not occur to me, but is that any different from the current approach? Since retirement is not penalized at present, this approach, which gives up any chance of winning, could be equally done now as in the socralynnek system, couldn't it?
It comes down to three considerations:
Do we want to incentivize completion? Current system does not.
Do we think that retiring to max score in face of certain defeat is somehow unsporting? Current system does not penalize this.
Do we think that retiring when winning due to running out of time should not be penalized? Current system does not penalize this.
If we like that, no change is required. If we think we would like to incentivize playing on without penalizing the time retirements, the socralynnek system seems to be a way to do this, if the scoring system is not too insane!
And if we think incentivizing completion is not a competition issue, then I am happy to leave it alone, not penalizing the time requirements, and allowing retirement to maintain score as an allowed exploit (posted as such, so all know). Then each player can decide whether to add such a farce score to their GR totals or not.
dV