I don't know which problem will be more difficult to solve. If the booster engines failed to light because of the speed it was falling, they could light some engines earlier to limit the speed it reaches which could be a simple software timing change.
If they lost attitude and roll control of the Starship, that might be a hardware problem which could take more time to engineer a solution.
why on Earth would the velocity that the Booster was falling have any bearing on the engine's re-lighting or not?
When it's in freefall, there's little accel /decel. Some moderate decel once it hits terminal velocity as it encounters thicker atmosphere. But these engines MUST re-light whatever the accel / decel parameters. As long as the fuel pumps can maintain CH4 and O2 pressures to all the necessary engines ....then they SHOULD relight. And it MIGHT simply be lack of good CH4/O2 pressures, we don't know what the exact tech issue is.
I am certain the Raptor engines reliability is of much bigger concern to the SpaceX engineers than attitude control and roll issues on Starship. SpaceX have done RCS thrusters for years and years, it probably needs some changes, sure, but that's fairly easy from engineering standpoint. All their Falcon9 and Falcon Heavy vehicles use RCS thrusters, as does Dragon Cargo and Dragon Crew.
Engines are an issue. SpaceX will absoluitely know it, but they won't talk much about it.
The mere fact that they chose a "RETURN TO EARTH TODAY IN a KNOWN REMOTE AREA OF THE INDIAN OCEAN GUARANTEED" suborbital trajectory (i.e. it's coming back today, whatever happens during flight) indicates they already know Engine relight is not reliable. Also they ditched the "Let's land within 60miles of Hawaii, yayyy!!" plan. No surprise, it's just a bad look if a SpaceX vehicle came careening in and parts crash-impacted near Hawaii.
EDIT for CLARITY: If a space vehicle is in ORBIT (not SUB-ORBIT that Starship flight was on) and the engines dismally fail to light during a de-orbit burn....then that's a PR disaster and potentially a crash-landing in a populated area as they have lost control of WHERE and exactly WHEN the vehicle will hit the Earth. Huge issue for SpaceX, very embarrassing and the FAA would have taken an *extremely* dim view of that.
A suborbital trajectory gives a KNOWN time and place for the vehicle to re-enter (whether the de-orbit burn happens or not) and so.....remnants hit the Southern Indian Ocean = safe for everyone (except a few fish) and no PR disaster.