Heterogenous high availability pairs - downsides?

Background/use case:
We had a B380 pair in HA mode.
Yesterday the master died, and the failover went smoothly.
However, while waiting for a replacement I am nervous about leaving us with only the one router in a critical piece of the infrastructure.
Therefore we added a B20X as the slave in a pair with the B380, with the B20X configured manually (I assume sync of the configuration from the master B380 should not work).
A quick test (pulling the plug on the B380) worked as it should, the B20X took over and life went on.
Powering up the B380 also worked well, it regained its mastery and the B20X returned to slave mode.
A lot of our configuration is performed using tags in IC2, so a need for manually keeping the B20X slave up to date is not too onerous

Question:
Given the above, what could go wrong? (Aside from fouling up configurations due to them being manually performed rather than automatic.)

Cheers,

Z

Configuration complexity may make this more onerous - different names for WAN interfaces used in OBP and so on being one quick thing that comes to mind.

My main concern in such a setup as a longer term solution would normally be around feature / performance parity in terms of the hardware used.

If your day to day network operation requires a Balance 380 using a lower specified unit could have unintended consequences if it were to become the primary whether due to a failure or some other issue.

I agree - definitely not a desirable set-up. We consider it (and for the short-term) as a way to minimize catastrophic failure if the remaining B380 goes belly-up - the local network and its resources would still be (mostly) available.

Cheers,

Z