"There is a pattern of 'fastest in the early part of the race, then even-out in the middle, then become slower than Mclaren - all due to the car not keeping tyres optimum as fuel load comes down' , and it has been consistently that way, on a variety of tracks of late (ever since Japan, I think).
This could be a fundamental baked-in problem of the RB20 and can't be 'solved once and for all' within the season, I think :"
There's another elemental dimension that could also promote this scenario.
If the engine is run at full mandatory maximum fuel flow some of the time, then sub maximum at particular periods of race to affect overall pace ....
Theres lucid discussion of this in review of McLaren MP4/4 ... with Honda turbo engine in 1988 ? Restricted by end of turbo era fuel capacity, they ran it with full "bleed" fuel allowance first part of race, knowing that they'd never get to the end. Qualifying on pole, wound up for first third of race, economy to the flag.
Extrapolate that to now, start race with notional 10Kg less fuel on board (projection, I don't know this) move away from competitors with lower weight (start weight with fuel etc doesn't have a minimum, as such) full power 1st third, diminishes as laps pass, target fuel usage comes down etc, leaving lack of options later in race.
Less weight at start, highest available fuel flow to begin, makes a powerful pace advantage. Depends if other teams look at this in such an extreme way, or run more distributed power "curve" for race distance. Can't really work if you start back in pack in regard to sub optimum fuel start weight.
Austria specific, less potential regen per lap, high gain from following competitors with DRS etc, starts looking vulnerable doesn't it ?