This is a fundamentally "what-if" scenario so it's not going to be perfect math, but we can look for hints. To go from non-GAAP to GAAP, you can add the adjustments up. The biggest non-GAAP adjustment (around 2/3rds of total) is stock-based compensation. That $401 million in stock-based compensation really can't be repeated, and wouldn't have been so large if they had IPO'd sooner. The next big they spent another ~$250million on driver incentives, almost entirely on UberEats.
How much of their expenses are required for a theoretical rides-only Uber? Impossible to say; they don't break out spending by segment. But we can make some guesses. The rides product has now existed for many, many years. Did they really need to become a ~22,000 employee company if all they were going to do is rides? I would guess no. So the question here is really: how much of their current expenses do we think only exist due non-rideshare business? The autonomy program is evidently an expensive bust. I see a lot of scooter marketing, which we can tell from this report isn't selling. Food is growing but they're in 3rd place, and it seems like they spend a lot of money to even get to third. So really the question is: are >30% of their expenses due to the non-ride business? I would guess yes.