Let me fill in the logic, then.
I make a spell to take an arrow / a spear / similar (a projectile) and make it travel along to hit the target. However, I use T: Part to only affect the shaft. Now we have a projectile following your spell design. It's a projectile, so by published ArM5 physics it should follow one of those three as there are specifically three. Which of the three does it follow? The only one that comes close is the first. So, this arrow/spear/similar is resisted. Now the big question. How is it that this is resisted but your spell is not???
As for Ghost Touch, what is the actual wording? There are cases in the rules where an effect is essentially T: Part but was written as T: Individual (cat's eyes, for instance). My books are in the box the bottom of a pile of boxes in a moving truck, so I can't check.
Here is maybe the strongest argument, though it's entirely RAI as opposed to RAW:
I could make basically the same spell for a projectile (as described above). As per the rules, it automatically hits. As per your ruling, it automatically penetrates. So anyone who uses the other methods must be an idiot. Why drop 1 magnitude to use Finesse against a fairly difficult value when you can hit automatically? Why keep a limit to damage because you're worried about penetration? Thus to me it seems like your rule makes no sense in the continuity of spells in canon and also violates the seeming requirement to either penetrate or hit via Finesse.
But I do understand how you're coming at it and strongly disagree with Jonathan on some points. In canon an animated corpse swinging a sword need not penetrate, just hit with an attack. I'm pretty sure a number of Jonathan's statements don't support that. Buffing a grog is fine, just don't buff his sword if you're worried about penetration.