So Ring 1 Stem is 96 beakers (3*32, from German play-through), apart from computing which is 127*3 = 380. You may randomly get a tech that you are half-way (or more) through researching. Since this is a straight 25 turn gig, it will be difficult to manipulate timing. Yes, you could have genetics by the end of the 25th turn, but that's not certain. While a fluke result (Adept Blue gives you Computing, which would be the best result) could certainly help greatly, I don't think it's that imbalanced, on average. No more than El Dorado as Spain, or Fountain of Youth.
It's a giant dice roll every 25 turns.
Computing wouldn't be open to start, but by the first 25 turn tech in most games one N2 and some N1L techs would be opened. (N0 being Habitation.) Say you research 1 N1 tech in the first 25 turns of trading, and have already gotten 1 N01 (Pioneering) to be able to trade at all of course.
That means you likely have tech opened like so: (beaker costs are estimates from the livestream, since we know those settings were where 25 turns was from)
1x N1L ~90
4x N2 ~100
1x-2x N2L ~250
0x-2x N3- ~400
1x-2x N3+ ~800
N3- are the cheap N3 techs, N3+ are the expensive N3 techs. Say we take Ecology, which opens up 2x N2L, 2x N3-, and 1x N3+:
1x N1L ~90
4x N2 ~100
2x N2L ~250
2x N3- ~400
1x N3+ ~800
That gives a total Science of 2590/10 = 259 expected Science from the first completion of the route. 800 is the max. 90 is the min. So this isn't "rarely" overpowered, it starts overpowered on average and just has an absurd upper bound. The +4 Science Trade Route in the same time would give 100 expected Science. You have a 1/10 chance that the max will hit, which is 8x the value of the other trade route.
So already the expected is far higher for Adept Blue on the first iteration and the variation is massive 90 to 800. This is a slots game mechanic, not a strategy game one.
It gets worse. Let's say you did get that 1/10 N3+ hit (Terraforming). Now you've opened up 2x N3+L, 1x N4+, and would be best served (from pure Science output standpoint) to research the N1L and/or N2L on the second iteration. Probably you can reasearch the N1L and 1 N2L. Which would put you at:
3x N2 ~100
2x N2L ~250
2x N3- ~400
2x N3+L ~1200
1x N4+ ~2600
That gives a total Science of 6600/10 = 660 expected Science from the second completion of the route. Even if we assume the other Trade Route would have made Level 2 after the first 25 turns of trading and that doubles the output, expected is now 3x the other Trade Routes. 2600 is the max, which is 13x (more likely 26x). 100 is the min... which is either balanced or not that great.
Now you're only going to hit the max max 1/100 times, but when you do hit that max you're talking a huge shift in Science output. You can grab a N4 tech in the first... ~75 turns?
If you're FI or SF you could actually grab 2x N4 techs (or a N4L, but that doesn't make as much sense) in that timeframe. Then the next iteration becomes really absurd.