I have yet to see any really solid arguments against either of those points.
In an autonomous-car world, there's going to be huge initial capital costs to buy the fleet.
Yes and no. It'll be easy to get debt to pay for something with an obviously high return on capital.
But it's local, only useful to few towns in Lithuania. Still worth taking a look at to see how simple and functional app should be done.
But arguably, Uber has paid the highest price by being first mover, allowing competition to follow behind and access this "social infrastructure" much more cheaply. As an example: Every Uber driver knows about Lyft (and other competitors) and is willing to drive for them if they pay better or offer more volume.
That said - there's no way for Uber to end up with a monopoly in this space. But if they did, they'd likely end up getting regulated like traditional taxi services.