AoA disagree does not imply airspeed disagree. Indicated airspeed is purely a pitot and static pressure driven measurement. AoA is there largely as part of the anti-stall warning system, e.g. driving the stick shaker/pusher. An important point of having AoA sensors is to provide a stall warning systems that is independent of IAS, you don't want these systems crossed with each other.

B737 airspeed disagree I believe just requires 5 knots disagreement for more than 5 seconds between left/right IAS. No AoA involved.

There may be some confusion happening due to the events in the Lion Air B737 Max. The maintenance crew tried to address what they believed were ADIRU/ADR issues in previous flights... including as a possible source of erratic airspeed indications and airspeed disagree warnings. So there are some discussions of airspeed disagree in relation to that aircraft. There may be a lot going on there we don't know about yet, with say faulty ADIRU/ADR causing multiple problems. The ADIRU has inputs from the AoA vane but does not use that in calculating airspeed, it is using that to drive stick shakers/pushers and optional display AoA, and unfortunately in these cases to also drive MCAS.

Having a AoA disagree indication (and/or full AoA display option) might have allowed more prompt diagnosis of problems, but only if pilots were aware of and trained on MCAS. And even then, yes it's a minor point. I was not intending to claim that itself is a deep solution here, just not having a disagree warning as standard was a bad decision. The entire system, and overall approach taken by Boeing seems tragically deeply flawed.

I think the Ethiopian preliminary report says they were in IAS disagree for the whole flight too. I don't know what caused the IAS disagree. It played a role in the crash, since its checklist disallowed lowering flaps (which would have disabled MCAS) or throttling back (which would have avoided surpassing Vmo, which likely made manual retrim impossible).

Ah now my addled brain finally wakes up, well maybe. What may be happening here is the ADIRU is doing AoA based static port compensation. It will normally be a small correction, and I was trying to keep that out of the discussion because I thought there there have been reports of erratic airspeed in previous Lion Air flights and that was not jiving with small changes of airspeed I was expecting (and now I'm going to reread the Lion Air preliminary report to check what was reported). But to start with you only need 5 knots to trip the IAS disagree, so sure I'll expect it could trigger that (and also possibly an altitude disagree). But potentially more intersting is what does that usually small static compensation change do when the ADIRU thinks the AoA is really out of whack, does it keep applying the "correction" or does it ignore it? So that question circles back to be similar to what Greenspun is asking in the original article.