I don't disagree with anything you wrote (and I'm well aware of these methodological issues, although I agree it's important to discuss them), but I am still having trouble with this particular judgment.
Let's compare another type of telicity, tangentially from your example above:
"I ate the hamburger"
Does that imply or assert that I ate the whole hamburger?
To me, in its basic usage, the meaning is telic: "I ate the hamburger in 5 minutes."
But I can also force a reading where it's partial consumption:
"I ate the hamburger, but not the whole thing."
Traditionally, we'd assume it is an implicature then.
And indeed we can force:
"I ate the hamburger for 5 minutes."
So does that mean this sentence is vague or ambiguous?
In some other languages (I think French, but also many others) you could indicate this distinction using a partitive. In English we don't have that. So again, is it vague or ambiguous? Is one basic and the other coerced by either syntactic or pragmatic context?
(Tangent: I've always been much more inclined than others to accept "coerced" readings as pragmatics, thus underspecified in semantics. Others seem to want more systematicity and rigidity in the semantics, so I don't agree with their narrow judgments. But an interesting follow-up to your reply would be to ask how to define "coercion", whether it's just a label for unintuitive cases, or if it's actually a proper technical notion between pragmatics and semantics, e.g. pragmatics bending semantics. If we allow coercion in a broad enough sense, we can basically bend all of semantics via pragmatics to the point where they two are no longer distinct, as you may be hinting in some comments above!)
To my ears, this is very different from an easily cancellable implicature like:
"The lone ranger got on his horse and rode into the sunset, but not necessarily in that order."
However, I should add that the hamburger example feels more natural in this way as well, while the "to/arrival" case seems even less easy to cancel. These are just very odd:
"He walked to the school, but he didn't arrive."
"He walked to the school, but he didn't get there."
"He walked to the school, but he stopped and went back home."
"He walked to the school, but it wasn't there."
I suppose some implicatures can be stronger than others. The simple methodological solution is to say that only absolutes are grammatical and leave it at that. But I do find there to be something interesting going on here, though I think I need to stop pondering this because my judgments aren't clear anymore.
By the way, there are several possibilities in this case including:
1. The specific meaning of "to" is up for interpretation: e.g. to the vicinity of something, to its center/edge?, to some relevant proximity?, (in the direction of?), to spatial location at some specific time*. This would allow for "success" to be relative to that meaning, and therefore harder to test by the methods we've been applying.
2. This sort of motion is ambiguous or otherwise under-specified, so the preposition is clear but the overall usage is not. This would mean we should test "to" in other sentences. "I threw the ball to you." seems clear enough in a context where I miss. So maybe it's more "proximity" than anything else, as in (1). Complicated!
[*Nominal tense is found in some languages, mirroring English "president-elect" and "ex-president", but are prepositions ever tensed in languages? I suppose in some languages some adpositions are basically verbs so they could be tensed somehow.]