We’re not ready for that. The missing pieces include lack of shielded support for some common uses of scripts (especially multisig and hash time-locked contracts), and the dependence of coinbase transactions on t-addresses.
That said, we could include some preparations in this upgrade, for example enabling payment of mining fees, block rewards, and maybe FR directly into Sapling addresses. There is also a possible method of supporting shielded multisig without a further consensus change, taking advantage of the fact that RedJubjub is a Schnorr-based signature scheme. Shielded HTLC support is more difficult because I think it requires a circuit change, but it probably wouldn’t be a blocker on its own for deprecating t-addresses.
If all goes well, I would expect that the next upgrade after Blossom, in 2020, would be a good candidate for disabling sending to t-addresses (but you could still send from them). This is purely my personal view; it is not something we’re committing to, and I haven’t discussed it with other Zcash engineers.