The purpose of letrozole is to control estrogen levels, with the typical goal being the low end of normal.
Tamoxifen (Nolvadex) and clomiphene (Clomid) and the SERMs in general don't control estrogen levels but do, in some tissues, block the action of estrogen. They also in other tissues can act as estrogens.
You bring up what would be a good subject to look into again. Quite some time back I did an exhaustive literature study on everything affecting LHRH and LH production, and I found reason to think that SERMs benefit LH production not only via anti-estrogenic effect but also via estrogenic effect in the hypothalamus or pituitary -- I'm afraid I don't recall which. That would bear looking into again.
Personally, I've tried using just letrozole post-cycle and did fine with it, but it seems I am at the better end of the range with regard to easy recovery from cycles. I consider the traditional SERM-based approach to be by far the most proven and continue to recommend it.
That said, controlling estrogen levels to low-normal post-cycle, if they otherwise would be, cannot be a bad thing. And so there's nothing unreasonable about combining the approaches, providing that the anti-aromatase dosage (e.g. letrozole) is not excessive. It definitely works fine to do that.