If your premisses is reasonable to assume, then it is also reasonable to assume--perhaps to the extend that the grammar allows it--to find the proposed connection reflected in PIE in a way that does not leave your assumption stand contradicted.
*-t is found in various instrumental morphemes [I mean, right?]. So our intuition does not count for much. As the ...
Yes, you are right
English Wiktionary tells you about the German word Bucht:
Via German Low German from Middle Low German bucht, from Old Saxon buht, from Proto-Germanic *buhtiz. Cognate with Dutch bocht.
German Wiktionary says:
von niederdeutsch bucht „Biegung, Krümmung“, „landeinwärts gebogene Strandlinie“, belegt seit ...
This sounds quite unlikely, because "biegen" is a strong verb, so the past participle ends with -en ("gebogen" in German, "bogen" in Old English). So it would be hard to explain the weak past participle ending -t.