Actually he mentioned other people who recommended 90%. In other threads he has mentioned that you can use rep quality-crisp reps, or RPE and that you may be 5-10% under printed program targets.
I use a recent rep based max estimate, but tend to program training loads that are 5-10% under what most programs recommend, so for me, 75% 5’s, 80% 3’s or 85% 2’s are “standard” heavy enough to produce strength gains. 80% 5s, 85% 3s and 90% 2’s are “heavy” and 70% 5’s, 75% 3’s and 80% 2’s are “light” or used for a quick neurological break. (I also adjust downward for lower body movements adjusting for body weight, but it usually ends up lowering percentages by about 5%). If I go in my heavy range all the time it is counterproductive. It is no more effective than “standard” and it is harder to recover from.
On a real rep max test, I assume I can get 85% for 5, 90% for 3 and 95% for 2 (adjusted for bodyweight on lower body exercises). While I can ‘survive’ that for quite a while, it is totally unnecessary.
Personally, I can’t use RPE and can’t always use rep quality because as I ramp, I can have the RPE DROP as weight increases in ranges close to my ideal training weight. Often if I repeat a weight twice, the second set feels easier and crisper. Sometimes I will hit a weight 10% under my goal training weight and it will feel heavier than expected, maybe a 8.5 RPE, then add 5% and its back down to an 8 for another jump or two, so I can’t self regulate, but since I do fine with strength gains using weights that are within 5-10% of my capacity, I don’t have to worry. Also, I don’t vary much from day to day. I can pretty much hit the same rep max in the same lift every day 7 days a week. I won’t get stronger but I can do it, so I have to use slightly reduced percentages to get the best progress.
I read Verkhovsky original works and he stressed that ALL training percentages such as the Prilepin chart were based on 90% of a competition weight and that lifters took weeks to recover after competition. Also the Russian research showed that even if an athlete can use moderate to heavy weights every day, even multiple times per day, the nervous system required lighter training days or it got stuck in sub-optimal firing patterns/rates and never improved.
Also, I have a theory about Westside max effort training, as well as using lift variations like the Russians did. I believe that using a max effort exercise that is different from your main lift is NOT a way to allow you to train super heavy, but rather it is a way to prevent you from training too heavy. When you pick up an exercise variation for the bench press for example, like a floor press with chains, you simply will not be neurologically efficient enough to over-stress yourself for 1-2 weeks. You may hit a max, but in reality it is only going to be around 90% of a max because you have not mastered the move. In comparison, a powerlifter will be so efficient in the regular bench press or squat or DL that a 100% lift will be extremely neurologically stressful. (actually I think that Westside in effect uses max effort for different purposes. Some exercises are full ROM replicating the main lift but with a different strength curve. These prevent over-stress due to novelty of strength curve. The second is max effort moves that target a reduced ROM or portion of a movement pattern like rack lockouts. These reduce overall stress by reducing ROM but do raise intensity in that ROM. In other words, they really use Max Effort days for two different purposes and will sometimes refer to ME exercises as “testers” and “builders”.