I'll chime in adding my support (my "single data point") that ST spells feel overly punished in BG3 compared to 5e.
- Some enemies in BG3 have inflated ability scores (e.g., Gith patrol), benefitting their saves more than AC.
- Some enemies in BG3 have reduced AC (e.g., goblins), making them easier to hit with attack-roll abilities but not ST spells.
- High ground in BG3 benefits attack-roll abilities only.
- Surface effects & grenades make it harder to keep concentration, and more concentration spells rely on enemies continually failing their STs than providing extra opportunity to hit enemies with spell attacks.

@Flooter, + other things relating to RNG in BG3
The sine wave pattern was removed with the addition of Weighted Dice in patch...4? Basically, Weighted Dice made it so a roll of 1-10 would more likely be followed by a roll 11-20 and vice versa, which in practice removes the large-scale sine effect.

However, 500+ recorded rolls using the weighted dice in patch 4 (5?) showed that some numbers (6 and 18) were still popping up a statistically significant anomalous amount of times.

I remember analyzing your 400 rolls, and found that the distribution was consistent with a normal RNG function. Specifically:
- the average was consistent with the expected average of 10.5 (I found 10.72 +/- 0.28)
- the distribution of rolls is consistent with uniform = each number appearing ~5% of the time (at the 95% confidence level, the chi-squared statistic of your results was 20, which is less than the critical "reject hypothesis" value of 30.14)
- the only number that appeared a noticeably different number of times was 5, 2.2-sigma less than the expected number of times. However, out of 20 numbers, it's perfectly reasonable for one to exhibit that degree of difference.
Flooter, do you remember if your recorded rolls were using Loaded Dice or not?