Don't have all the details on the beverage preparation as I don't have access to the paper, but from the summary below I infer that 1 gram of carbohydrates per kg per hour (ie. if you are 70kg, you would need 70gr of carbohydrates per each of the four hours = 280gr of carbs) in a 2250ml (almost 4 pints!!!) mix of cocoa, any of the mentioned milks and water should work.
I think this could be applicable for the first day at L'etape de la Defonce (11k TTT followed around three hours later by a 85k stage) so I'm going to try something similar as a recovery drink after some turbo sessions in the evening and see how I feel riding hard on the following morning.
Cycling Time Trial Performance 4 Hours After Glycogen-Lowering Exercise Is Similarly Enhanced by Recovery Nondairy Chocolate Beverages Versus Chocolate Milk
2016, 26, 65 – 70
Postexercise chocolate milk ingestion has been shown to enhance both glycogen resynthesis and subsequent exercise performance. To assess whether nondairy chocolate beverage ingestion post–glycogen-lowering exercise can enhance 20-km cycling time trial performance 4 hr later, eight healthy trained male cyclists (21.8 ± 2.3y, VO2max = 61.2 ± 1.4 ml·kg-1·min-1; M ± SD) completed a series of intense cycling intervals designed to lower muscle glycogen (Jentjens & Jeukendrup, 2003) followed by 4 hr of recovery and a subsequent 20-km cycling time trial. During the first 2 hr of recovery, participants ingested chocolate dairy milk (DAIRYCHOC), chocolate soy beverage (SOYCHOC), chocolate hemp beverage (HEMPCHOC), low-fat dairy milk (MILK), or a low-energy artificially sweetened, flavored beverage (PLACEBO) at 30-min intervals in a double-blind, counterbalanced repeated-measures design. All drinks, except the PLACEBO (247 kJ) were isoenergetic (2,107 kJ), and all chocolate-flavored drinks provided 1-g CHO·kg body mass-1·h-1. Fluid intake across treatments was equalized (2,262 ± 148 ml) by ingesting appropriate quantities of water based on drink intake. The CHO:PRO ratio was 4:1, 1.5:1, 4:1, and 6:1 for DAIRYCHOC, MILK, SOYCHOC, and HEMPCHOC, respectively. One-way analysis of variance with repeated measures showed time trial performance (DAIRYCHOC = 34.58 ± 2.5 min, SOYCHOC = 34.83 ± 2.2 min, HEMPCHOC = 34.88 ± 1.1 min, MILK = 34.47 ± 1.7 min) was enhanced similarly vs PLACEBO (37.85 ± 2.1) for all treatments (p = .019) These data suggest that postexercise macronutrient and total energy intake are more important for same-day 20-km cycling time trial performance after glycogen-lowering exercise than protein type or protein-to-carbohydrate ratio.