[A belated reply to this original question before I launch on Amman Iar
and the gardenpath David's been led up by our colleagues Matt & Dirk ;-)]

Barry asks:
> I've always wondered how many other conlangs besides mine lay importance
> on stress to indicate the correct meanings of words? I found it a nifty
> feature of  Tagalog when I studied it because it seemed to allow you to
> stretch out the lexicon.

In Livagian, stress (= nonlow tone) is on the first syllable (simplifying
somewhat) for two reasons. First, it is an ingredient in achieving that
loglangy goal of having unambiguous word boundaries. Second, the
psychologically most salient parts of the phonological word are its start
and its primary-stressed portion; I wished for these parts to be congruent
rather than in conflict.

But I also share your sense that it serves to stretch out the lexicon
(for words of a given length). For that reason, among others, I allow
the phonological forms of lexical items to consist of a sequence of
more than one phonological word. This then expands the number of
possible tunes that the entire phonological form can have.

BTW, my primary muse for conlanging is and has always been English. I seem to
be the only conlanger over the age of 12 for whom this is the case. So where
you identify mobile stress as a nifty feature of Tagalog, I would identify it
as a nifty feature of English. I am in the process of doing an overhaul
of Livagian phonology & was utterly delighted to find that the consonantal
phoneme inventory looks basically like English's with a few gaps filled in
the velar series and with a few clicks flung in. I've always felt that
English gets unjustly slighted by (anglophone) conlangers.