In the long run, yes.

How do you account for tribological degradation in tolerances over time? There are techniques to negate the problem entirely, my 40+ year old Japanese motorcycle uses high-ish pressure hydrodynamic bearings to allow oil to be the "wear" surface... But it isn't foolproof.

Newer techniques rely on better metallurgy, oil additives which precipitate on designated surfaces, etc etc etc. It's one of those things you and I have both observed where the more you look at it, the more complex it gets.

Now change gears (haha) and observe the complexity of analog/digital systems. You and I front-load the complexity, so that any configuration changes should (mostly) be possible in software. "Oh no, we didn't use this particular junction of p-n MOSFETs"... They added $3 to the BOM, but our flexibility/redundancy is increased.

Look, you should know as well as any EE that solid state systems, when run within parameters, have a tremendous unchanging mtbf. While there are currently issues with undersizing MOSFETs and igbts, I expect our engineering models (in our heads) of what to expect from these evolving electronic components to adapt to these new operating environs, the already reasonable reliability will only increase.

Failure of these packs and their bms' are low enough that we can point to specific failures, versus how often we see someone with serious ICE engine issues in the local mechanic's shop.