Whoa hoss. Sikaflex 295UV is paintable and sandable, and probably bondable. However, the structure proposed here makes no sense, and the result is predictable, as Scroll describes.
You've bedded the plexiglass in an elastomer rather than flox/epoxy. In essence, you're floating the plexiglass on a 35 Shore A jello base. Nothing wrong with that idea. In fact, I think it to be an excellent choice in some applications, notably the long RV-8 rear canopy-to-frame joint.
However, if I understand correctly, here you're then trying to add a much stiffer epoxy/glass overlay connecting the aluminum to the plexiglass (and the sika) via an epoxy surface bond. There are only two possible outcomes. One, the plexiglass will move around (it has a huge coefficient of thermal expansion) and either crack or debond the epoxy/glass. Or, two, the epoxy/glass layup can be made thick enough/strong enough (including enough bond area) to resist the forces from the plexiglass and aluminum expansion/contraction rates...in which case the sikaflex was pointless. At best, it has been relegated to being a sealant and nothing more. The epoxy/glass is now the structural member.
If the windshield is bonded with Sikaflex, the fairing at the base of the windshield should be built as a stand-alone component, and subsequently be bonded into place with more Sikaflex, so it too can float in relation to the other parts.
Slick-tape the entire fairing area, mask everything else, shoot a little PVA, then do a fairing layup. When cured, peel it off, trim, sand, and finish the surface, much as we might finish the empennage fairing. I'd take it all the way to finish paint color. Then lay it into place, run 3M Fineline tape on the plexiglass a consistent distance from the edge (pick something, 1/16" to 1/8" should look right), and mask back from there. Sika the fairing into place and wipe all the squeeze-out to tape thickness along the tape, which when peeled, defines a nice finished edge.
