It will be interesting to see how they are planning to earn money. The only ones willing to expend resources on space cleanup right now seem to be ESA (the European Space Agency).
Lobbying for regulation mandating a service you provide is a pretty tried and tested business model. Every company IT audit I’ve ever had the pleasure of participating in works the same way.
"Space-tug" like services from a third-party would be vastly more expensive. The only use case would to de-orbit a satellite that for some reason failed to de-orbit itself.
It's not "just" some delta-v, it's the same amount of delta-v as the original perigee kick. Most satellites get their perigee kick from a booster stage from their launch vehicle that separates from the satellite and de-orbits.
Every gram of mass of fuel and engine on a satellite is a gram less of payload.
That's pretty much most of the world's nations' telecommunication needs. SpaceX has put up at least half a dozen satellites for small nations meeting these very common requirements.
I have no idea how much force it takes to actually move a given amount of mass out of orbit, and I assume it is different depending on the height of that orbit. So maybe doing this would simply add too much weight in the form of fuel storage to be feasible. In my head though, orbits are fragile and it shouldn't take a lot to nudge something into a death spiral towards the atmosphere.
In addition, the bulk of the space-junk problem is not defunct satellites, but fragments from previous collisions, and stage-separations (screws, scraps, paint-chips, etc). Satellites (even defunct ones) are easily trackable, and have known trajectories. Random 10 cm pieces of metal are not.
While mandating EOL maneuvers for low-earth satellites is definitely a solution to part of the problem, it is not the full solution.
It takes a ton (possibly literally) of propellant to change your orbit from a circle in GEO to an ellipse that intercepts the atmosphere. If you don't have enough, you've not put it in a deorbiting spiral, simply an oval that will stay up for about as long as the original satellite's circular orbit.
Rocket scientists work in terms of 'delta V', meaning an amount of energy gives you a particular change in velocity. GEO to LEO takes on the order of 6 km/s (oversimplifying because orbital mechanics are complicated), meaning you need a rocket engine that's capable of accelerating your satellite from 0 to 13,400 miles per hour.
That's really expensive and difficult.
GEO is likewise not a problem because things out there are really far apart and they generally don't cross each other's orbits.
MEO (where GNSS services live) is the real danger. Thankfully there aren't a ton of objects there, but deorbiting the dead ones should be a very high priority on anyone's list.
SpaceX satellites are passively de-conflicted. Which means that the orbits are planned such that the satellites never can be at the same place at the same point in time. But yes SpaceX does care a lot about debris in their orbital shell.