Have they been working on this algorithm before
No, they just made a list of algorithms, and picked it "randomly", by
checking testnet4 block headers. Of course, testnet4 is quite bad choice, because of the 20 minute rule, and you can use it, to control the outcome, but anyway: their chain, their choices.
Because that's how slow such changes must be.
Well, when they announced BIP-110, they could see miners signalling for months. They had a chance to see months in advance, how many miners will support their rules. And they did nothing with it, by just assuming, that miners will suddenly support it at the last moment, just because they wish it to be true. However, they didn't think, that if it would be the case, then all months of signalling in advance would be pointless, and everything could be handled as "please fork at this block number" instead.
It is not about being bitcoin or altcoins either.
It kinda is. Altcoin developers like hard-forks, and changing rules on the fly, because they have a small market cap, and you can ask two or three peers. If they all agree, then you can introduce new rules tomorrow. That's how altcoins operate. They are just a playground for more serious projects, where we can see exactly, what would happen, if we would change this or that rule.
eg. testnet6
We don't have testnet5 yet. And instead of releasing it, people are
waiting for fjahr to do that, because people don't care about having a working testnet here and now. They care about it being included into Bitcoin Core, to have a bigger starting price, and sell it for BTCs, when AltQuick will list it.
If they didn't take all those steps then this is yet another reason for us to call BLuke a shitcoin.
Why should they care about deploying things properly, if they didn't care earlier, when they could trace months in advance, who is against BIP-110, and who supports it? They had a lot of feedback, probably more, than they deserve. A lot of attention from Core developers, and not only them. A lot of interviews, podcasts, forum topics, and discussions. And they ignored all of that feedback, and deployed it anyway, in a form, that was destined to be rejected by the majority.
And even when they forked, exactly as they wanted, then it was still not the end of the world. They could still gradually decrease their difficulty, by introducing weak blocks, fractional confirmations, and keep improving their chain in a soft-forked way, by having a slow, and stable development, just like you described. More than that: they could also remove some spam, without making their nodes incompatible, and without introducing any consensus changes, and any kind of soft-fork. There are
some implementations, doing it in that way, but BIP-110 supporters just decided, to reject that path, even though they were informed about it.
So, if you consider all of that, then it can be simply summed up as: they got, what they wanted. And now, they have their own coin, with their own rules, and with Knots, being the "reference implementation". Also, it doesn't stop to surprise me, that they blocked P2PK, which was harmless, used by Satoshi, and consumes less on-chain bytes, than other address types, because public key hashes are never pushed on-chain, and it is guaranteed by consensus rules to be spent by a single signature (unlike Taproot, which can include big TapScripts).
