Thanks, narhiril! You're awesome.
These posts make my day. Seriously. Thanks.
I've been playing a lot today (something I haven't had time for in a long time) and I'm realizing that the gap between tiers 2 and 3 of the tech tree isn't behaving as reliably as it used to due to the new experiment result system.
Consequently, I'm spending tonight tweaking some percentages to try to make examination of experiment results more reliable. I had underestimated the amount of input required to keep the reaction running smoothly, and it's resulting in a lot more micromanagement and frustration than the old system. I'm aiming to remedy that with the next patch.
I'm starting by sharply increasing the percentages of all tech advancement unlocks - some by a few points, some by an order of magnitude. I'm also removing the distinctions between them - one advance will no longer be more common than others. This worked fairly well on the "slab of accreditation" system, where it was perfectly reasonable to have one or two training workshops feed a documentation reaction indefinitely, but it isn't working so well with the advances being tied to experiments - which require actual reagents. The evaluation reaction is catching up to the experiment progress, even when the experiments are given a substantial head start, causing the player to have to constantly monitor it to look for advances which are nowhere near as common as they should be.
In short, it's getting an overhaul this week. Evaluation will more reliably unlock new advances, and the evaluation reaction will queue itself whenever reagents are available for it. In addition, experiment results should not have quality levels - this was an oversight on my part that I am taking this opportunity to resolve as well.