Due to time dilation, an outside observer never sees a falling object actually cross the event horizon.
This is correct, but perhaps not quite in the way that you visualize it.
It's trivially true, in a literal sense, that an outside observer can never see anything cross the event horizon, since this would require seeing inside the event horizon, and no light (or other information) can escape from there to the outside. The most the outside observer could hope to see, even in principle, would be the object hitting the event horizon and disappearing (but they can't even quite see that, for the reasons discussed below).
It's also true that, the closer the falling object gets to the event horizon, the longer it will take for any light it emits to get away from the black hole (and the less of it will make it out at all, and the more redshifted the light that does escape will be). In fact, as the object approaches the event horizon, the time it takes for light emitted by it to escape the black hole tends to infinity.
However, trying to interpret and visualize this as:
[an] effect time dilation has on objects near the event horizon, causing them to slow down to a halt from the frame of reference of an outside observer
is probably not the best way to think about it. A better way of looking at what's happening is that time dilation is causing light (and other signals) to take longer to escape the black hole, and thus the outside observer is seeing things that happened a longer time ago in the past.
In other words, the object falling into the black hole isn't slowing down. As benrg notes, it actually speeds up quickly and falls through the event horizon in finite time. And in doing so it adds mass to the black hole and causes the event horizon to grow bigger.
All that's slowing down (from the outside observer's perspective) is the light that the falling object emits as it's falling in. So the outside observer can in fact "see" the black hole form and grow, at least as much as it's meaningful to speak of seeing a black hole at all. (They obviously cannot receive any light or other information directly from the black hole itself. All they can observe is the infalling mass and the extreme spacetime curvature and its associated phenomena that are consistent with there being a black hole in the middle.) But the outside observer is also seeing (very small amounts of very strongly redshifted) light emitted by objects that fell into the black hole a very long time ago, including some emitted by objects that fell in when the black hole (and specifically its event horizon) was smaller than it is now.
Interpreting these observations in a particular way, the outside observer could claim that they're seeing light coming from inside the black hole's current event horizon (as calculated based on indirect observations). But there's no actual contradiction there, since that light was emitted back when the event horizon didn't extend that far yet!
Of course, in practice, all of this is mostly an exercise in pedantry and deliberately quirky interpretation of observations. What a real outside observer actually sees, as you correctly note, is objects fading away as they fall into the black hole, and the approximate place where they fade away (accounting for the massive optical distortion around the black hole) being approximately where they can calculate that the event horizon should be.
All the rest is just theory, and not particularly fruitful theory either. Sure, you can say that in theory you could sit in a spacecraft orbiting a black hole and see light emitted by an object that fell into it a million years ago, because it's theoretically possible for a photon emitted by the object very close to the event horizon to take a million years to reach you. But in practice you just can't, no matter what kind of incredible sensor technology you may postulate you'd have, because the probability that you'll ever detect a photon emitted during that vanishingly small portion of the object's descent is for all practical purposes zero.