By stable, I meant something that doesn't have the unpredictable UI variations of checkout.js, which for some of us is a significant concern. (See various previous HN discussions about Stripe, particularly around the time the phone number/remember me changes came in.)
When there was always the alternative of using stripe.js (as also repeatedly mentioned in previous HN discussions) this wasn't a huge deal, because anyone who wanted stability could just go the other path. I think some of the concerns before were more about how many of us didn't realise that Stripe sees checkout.js as an ever-evolving test bed.
If that no longer applies then there is now a need to choose between instability with checkout.js and losing features with stripe.js. When I first joined this discussion today, I was happy at the idea of Stripe starting to support a broader range of payment options. As I'm reading more, my joy is turning to concern that checkout.js is becoming the preferred integration method with significant downsides for those who don't comply.
I do understand that there may be good reasons for locking in your choice of payment UI. On the other hand, this is not the first time I've dealt with a payment service where you can't cleanly integrate the whole process into your own brand/UX, and interrupting that flow rarely turns out happily for conversion rates or, sometimes, brand reputation.
With that in mind, if proper support isn't available for Alipay, then some sort of alipay.js as you mentioned would certainly be welcome, particularly if it allows at least basic styling customisation to match the colour scheme and typography of the merchant's site.