Expected score from an excellent combat would be
(Combat expertise x 13)Madness Multiplier)
(7 x 13) 13) = 1183
Expected score from a great combat would be
(5 x 13) 13) = 845

1183 - 845 = the score loss from going to turn 2 is 338 total.

The statement made is “Taking the scarab, and turn 2 win is less score than winning on turn 1.” Here is the proof.

The image on the left is the score with a turn 1 kill.

The image on the right has the turn 2 kill with the scarab.

Notably, the T2 with scarab has a score gain over the T1 kill. But this gain will not remain forever.

The score differential comes from two different factors. Combat Expertise and XP gained.

We’re it that XP gains were linear with score gains, this would be a cut and finished problem, but we have to account for XP Clamping when doing score math.

111,774 - 110,812 gives us a score differential of 962

1300 - 338 = 962
So we can confirm the score differential

Now let’s account for clamping. If we assume the maximum possible score retention from clamping the scarab, we come to the following:
50 xp each
/10 for the act 3 clamp = 5 xp each
/10 for the act 4 clamp = 0.5 xp each
Round up to 1 xp each, then combine to 4 total XP.

ScoreXP = ((experience/2)Madness Multiplier)
ScoreXP = (4/2)13) = 26

So we can conclude that the MOST score we could achieve from a scarab in this situation, once we have accounted for act clamping, is 26, which is less than the 338 we would get for clearing the combat on turn 1

Therefore, it is a score loss to go to turn 2 for a scarab.

A scarab gives 50 XP to each character, totalling 200. At this point of the game, we are PRE CLAMP.
This gives a score gain of:
((experience/2)Madness Multiplier)
((200/2)x13) = 1300 score gained from killing the scarab