They are likely to pay higher fees compared to other services/wallets that have direct channels to the destinations they want to reach. To pay a CashApp user it's also very likely that you have to pay more compared to other services.

The c= node adds an extra hop to every route that goes from/to the CashApp nodes. Fees c= charges are 1000 to 3000 ppm, so 0.1 to 0.3 percent is added to each payment.

Since most of the volume CashApp produces is sending, the c= node gets outbound heavy so they have to close the channels that the CashApp nodes open. They use the funds from those closed channels to open new channels to the destinations the CashApp users want to reach. This is how they are able to earn so much from "routing".

Originally CashApp did only peer with select nodes because of legal and compliance reasons. I guess at some point they figured they could earn a bunch if they put a node running LDK between them and the rest of the network. It's frankly a bit strange how they act like they earn from routing while they are just an exclusive gatekeeper for CashApp lightning.

This has nothing to do with the sort of routing that pretty much any other node in the network is doing.

The c= node has an exclusivity deal with CashApp so all their users's payment have to go through c=.

If the CashApp nodes would be open to accept channels from anyone the c= APR would be much lower because their entire operation depends on the special setup.

It's basically a combination of an accounting trick combined with obfuscated fees for the CashApp users.

CashApp could just charge their users a spending fee but instead they do not and split it of to c= which then claims to earn by "routing".

Nice, but the rate will have to increase to make it attractive. It's easy to find ways to earn more than 1500ppm and those alternatives have other benefits on top op them (like it doesn't fill up my boltz channel and with opening a sink channel I can rebalance it for future profits).

I would be interested in this when rates were somewhere towards 3000ppm.

Yes it can be supported by a future firmware upgrade (which is in fact planned for a while now).

The clients I've used all support taproot but the device can't handle it yet.

Bought two of them a while ago but I soon found out they do not support taproot addresses which makes them useless for my intended use. Cool devices though.

Ah yes for sure. Totally forgot that live-action movie even existed.

Ghost in the Shell

Be careful though, if you have a mixed fee structure (serving both cheap rebalance txs and high priority/time_pref txs) then it's hard for your peers to price you and your volume might suffer.

I understand the "21 million forever" concept but that's not my question or the original question (I think)

What do you specifically think about the idea of a fixed reward that Peter Todd recently opted for discussion on the bitcoin-dev mailing list?

In short: organically coins/sats are lost and therefore a small inflationary reward for miners might not absolutely impact the 21 million cap.

Point well made and taken. Too early to think about it.

I was hoping for a more nuanced response from @softsimon since he is more aware of the fee market development and likely has thought more about it therefore.

How do you know most bitcoiners disagree? Is it subjective or do you have some data?

I've been following the fee market development closely and I am not that optimistic. Once I thought that introducing an inflationary component to bitcoin was heresy but the past few years made me question that.

What Peter Todd suggests does seem to make sense considering the coins that are lost in organic ways.

Bitcoin is becoming increasingly important to the energy sector and having a tiny block rewards without any meaningful fees will impacts that progress.