On the one hand, this is technically in everyone's best interests not to do this kind of thing. The reason is, a successful intercept leaves a lot of debris scattered in an orbit, still travelling at orbital velocities. Meaning, it isn't coming down for a long time. It is already getting crowded and dangerous up there at some altitudes. When the Missile Defense Agency conducts tests they are always (with one exception) against sub-orbital targets. This means the debris lacks orbital velocity/energy and is coming right back down.
A couple of reasons you might want to do this is it is a natural progression or step up from what we're already doing. Generally in development of any new technology you want to take baby steps. You want to build on knowledge gained. You do not, generally, want to make giant leaps into the unknown. So an incremental step up from hitting sub-orbital objects to orbital makes sense. After all, it is more challenging. The primary differences being you need more energy, more capability in your interceptor to get further up there. You also need better sensor/control/guidance systems in your kill vehicle because the closing velocities are going to be higher. You need sensors that can see further and with higher accuracy so you have more time to adjust terminal guidance. You need faster computers and algorithms to "close the tracking loop" to an intercept solution. You need better guidance systems that can make both fine early adjustments and possibly large delta-V at the very end of terminal guidance. So with all these improved systems with upgraded capabilities from your sub-orbital (ie. ballistic) interceptors you'd want to incrementally test these.
Of course by specifying "direct ascent" they are leaving the door open to satellite based systems. One of the concerns there is that you have basically the same problems with testing interceptors and the same results: debris at orbital altitudes & velocities. The alternative is actually more concerning. If you arm your anti-satellite satellites with lasers then you back your potential adversaries into a very dangerous corner. Suppose you know XYZ satellite has a laser on it capable of melting your solar panels, overloading and burning out your optical sensors. Tensions are running high between your country and whoever controls XYZ. You get an alert that XYZ is now both close to and oriented towards one of your primary satellites... Unlike ballistic missiles or bombers where decision makers have minutes or even hours to contemplate options, your satellite could be dead any second, you have to act, now. This could lead to some rash decisions and escalations.
So while the decision to not perform this kind of testing makes sense. Much of the necessary technology is already developed and can be tested individually or together in simulations... It does push things in a potentially bad direction. Plus I have an inherent distrust for anything advocated by 'rats. 99.999% of the time whatever they want to do, just the opposite is what is probably best for the US.
I think the bigger issue that was telegraphed by this announcement was that Russia and the US are moving into space as part of this conflict. We and the Russians have always maintained a space combat environment up there.
One possible solution to the debris issue would be disabling without destruction, perhaps with directed energy and/or EMP. Turn a functioning satellite into dead metal. Whatever the Nut and the Slut may announce, the private defense industry isn’t going to stop working on this.