If the goal is to start the d-rest when the fermentation is at least 75% complete,
I've never heard it stated that way. It is usually asserted that one starts the diacetyl rest or temperature glide when there are 1.5 to 2 ° of extract remaining i.e. when the AE is 1.5 - 2 ° (6 - 8 points) above the expected terminal AE. But how does one know what terminal is? One way to determine when you are a couple of degrees above terminal is to do a forced fermentation test in which you remove 100 mL if fermenting beer and put it in a hydrometer test jar which you remove to a warm place. This will ferment out faster than the colder wort in the main fermenter. You then measure the AE of that and assume that the beer will eventually come to that AE and start your glide or diacetyl rest when the AE of the beer is a couple of degrees above the test AE. A more experienced brewer has brewed the beer in question or one very similar to it many times before and knows by other signs such as gas evolution rate when it is time to start temp. down. I my brewing, for example, when gas evolution slows I spund the fermenter and then monitor gas pressure. When it stops rising it is time to start the temperature down (I don't do diacetyl rests - the gradual cooling produces a better, more stable beer).
then unless I'm missing something, I'm not concerned about "total" apparent attenuation.
I think you are. In the examples you gave it is analysis of the AE that lets us detect that you are experiencing weak fermentations for some reason.
.
I guess a second way of describing it is that your method suggests a d-rest when the beer is essentially done fermenting. That's just never been my understanding.
As noted above I don't suggest diacetyl rests. That method was developed by commercial brewers who are looking for a quick way to get the beer off the loading dock. The traditional methods of cool fermentation followed by a gradual decrease in temperature to near freezing and long lagering make, IMO, better beer (with no diacetyl beyond the threshold amount which is an important part of Bohemian Pilsner).
But in any case I am not suggesting that the temperature glide, or diacetyl rest be done when the beer is finished fermenting. I am suggesting that they commence when the AE is within 2 ° of terminal. I was not, in the previous posts, trying to tell you when to start your diacetyl rest but rather using AE to diagnose your fermentation problem.
As for picking 1.020, I was simply going for a nice mid point in the style that would give me malty, but not particularly sweet. I know I was pushing the bounds of the yeast's typical performance but thought I might get there with a big starter and a pretty fresh vial.
Your approach has been to assume that you will get
20 points from a 90 point wort when the manufacturers specs tell you will get between 22 and 27. This is not a good assumption on which to base your decision to start a diacetyl rest because you do not know that you are going to get 20 - in fact it is very unlikely as you have seen. You, in fact, got less attenuation at 33 than the lower expected limit. A more reasonable approach would be to take the midrange specified by the manufacturer which would give you 25 points and plan your rest for 2 ° above that or 33 which is what you wound up doing but it did not work for you because your attenuation was not even as much as the manufacturers specified minimum. You must fix your fermentation problem.
A forced fermentation sample or monitoring of the gas evolution or plotting of the AE on a daily basis would have indicated to you that you weren't getting the attenuation you should. You might then have taken action such as repitching more yeast or adding vigorously fermenting wort (krausening). That might save these beers at this point and will certainly remove any diacetyl.