Let's talk about the elephant in the room: The Storj economic model (node operator payout model)

I like the idea of tuning the Reed Solomon settings. Though not for the same reasons. Even if Storj would over lower redundancy tiers, it would look bad reputation wise if they lost data. Furthermore, having fewer pieces of redundancy would just trigger more costly repair.

However, lower expansion rates could possibly be created by raising both the Reed Solomon numbers. So instead of 29/80, they could go 100/160 (just a random example I didn’t actually calculate what the impact would be). Something like that could actually offer similar reliability with the only downside of having more overhead, which could be partially fixed by increasing the segment size to 128MB as well. This could ensure similar reliability at the cost of not dealing as well with smaller file sizes. And it would incur a higher segment fee as well due to higher cost of metadata storage for more pieces to track. But if customers use the max segment size, it would match total cost for segment fees because you would also have half the amount of total segments.

I believe something like this has been mentioned by Storj Labs in the past as well. So this may already be considered.