Cases where the challenge come from the donor don't seem to fall under the scope of the OP's strategy: a donor's challenges would be evaluated on benefit to the donor, maybe in terms of marketing their brand to the community, as opposed to how it affects donations. As you use the word "genuine" for that case, maybe the idea of whether matching grants are "effective" or not is simply an incorrect question?
I personally am aware of more matching challenges and capping grants that are as I say "genuine".
You make a good point, perhaps their efficacy is more aptly measured on the outcomes of the major donor. However, it's still a commonly held belief in resource development and fundraising that a match is a very strong motivator to give. This is a good start to determining how true that is.
The OPs suggestion of heavily studying your own matching strategy and running different types is good advice (though maybe not 100% practical).
I'll also note that I couldn't find any particular data on the efficacy of a match in large campaign fundraising vs. in-room fundraising. it's very different to see a match in a direct mail or general campaign, vs. say a featured guest at a fundraising party challenging the room personally.
