Proposing an un-housed connection for the knee brace at both ends, and then a arrangement with an elongated peg mortise at the brace's lower tenon, along with a deepened mortise to allows the knee brace to, well, NOT perform its function.
Firstly, making the connection un-housed weakens the connection greatly - a brace works most effectively/directly if the end grain of the brace can bear more or less upon end grain in the receiving member, and if as much of the fiber in the brace can be directed into the receiving member - and that means housings in the post and beam to encapsulate as much of the brace section as possible.
As the engineer noted, "The downside of this is that the brace will offer no lateral stability to the exposed frame during raising. You'll have to make sure the frame is well guyed and temporarily braced until the shear walls are constructed."
Which begs the question of why would one bother to mortise and tenon braces in to place if they are to have no structural function in such a configuration as the poster has suggested? Even the modest 'function' of braces to hold parts together, like a beam to a post, during raising is defeated in this arrangement. The shear walls (I presume he means SIPs or interior walls with other forms of shear components) do the work after the frame is together, which again renders the braces superfluous. Scabbing on temporary braces such as 1x4's, as the poster mentions, makes more sense in terms of the final structural arrangement where other shear components do the work. In the arrangement suggested by the engineer, the knee braces become essentially decorative.
Knee brace 'function' when loaded, as the original poster noted, does involve creating high tension and prying loads at the beam's connection atop the post (or at the beam's end joint at the side of the post if the post extends further up beyond the connecting beam). Perhaps the problem is more the knee brace than anything else? Perhaps a knee brace sized so that it bends a certain amount when loaded, to absorb some of those tension/prying loads would be a better option? Or no knee brace at all.
Finally, the relish on the end of the brace tenon is generally rather minimal and prone to failure (especially if subject to tension loads), and some expectation of the connection maintaining integrity over the years as that tenon end portion rolls/slides over the peg in the post seems rather optimistic.
The 'solution' of creating a knee brace connection which results in a brace which cannot function as a brace is supposed to, however, does not make much sense to me. Perhaps there is more to this than meets the eye - a diagram might help make the picture clearer.
