Consensus is important for both taproot and separately for the activation mechanism. There are more soft-forks that Bitcoin will need, so it is important to achieve positive progress on the activation topic also, not get impatient and rush something ill-considered. Not all future soft-forks maybe as widely supported as taproot, and yet it could become existentially critical that Bitcoin prevails in achieving a future upgrade in dramatic circumstances, even against powerful interests counter to Bitcoin user and investors interests. We should treat the activation topic in a considered way and with decorum, provide tight non-emotive reasoning devoid of frustration and impatience. This is a low drama and convenient time to incrementally improve activation. People have varied views about the deciding factor, or even which factors resulted in segwit activating after BIP 141 failed using BIP 9. We do not have to solve everything in one step, incremental improvement is good, for complex unintuitive topics, to learn as we go - and it should not be hard to do less badly than what transpired leading up to BIP 148 and BIP 91. Failure to upgrade if permanent, or demoralizing to protocol researchers could be a systemic risk in itself as there are more upgrades Bitcoin will need. We are not Ents but we should use our collective ingenuity to find an incremental improvement for activation.
Great high level thoughts
The Ents themselves were created in Tolkien's fork of Shakespeare, when he was frustrated to learn that trees didnt actually march :)
Having followed standards for 10+ years consensus can be tricky
IIRC last time with segwit there was a straw poll in the wiki where devs could express leanings in an informal, async way. Something like that could be of value.
There's an insightful spec written at the IETF "On Consensus and Humming in the IETF", then IMHO is worth reading
That said, if we could find an incorruptible machine that could gather the highest fee tx from the mempool and post it every 10 minutes, bitcoin would largely run itself. So, while understanding the gravity of each change, we could perhaps have the mindset that there are a finite number, such that when complete bitcoin will move to an endgame where for the user it 'just works', much like the internet. If devs and changes are needed less, that could be viewed as a sign of success. This is a hand wavy way of saying that forks could potentially be a diminishing issue over time
Just my 2 satoshis