We always love to hear the ideas that people have -- part of why we're so successful at keeping the site running is because we have a really good match between the paid account features and what people want out of their accounts, and that's taken a lot of user feedback to get right.
(Before you spend time really thinking things through, though, you should bear in mind that there are a lot of commonly-proposed ideas, such as "let people buy icon slots without having a paid account" and "let people buy access to specific paid features individually for smaller amounts in the long run" that we don't do because it would actually make things worse in the long run. The mailing list messages where I explained the account system and account costs at exhaustive length were, unfortunately, eaten by a case of "we thought we were backing up that old mailing list server and oh fuck, we weren't", but I can do the explanation again if you want it. The short version: we set the prices of paid accounts by what we projected it would cost us to provide the service for everyone, divided by the percentage of accounts that we thought would be willing to pay us, so it's not "each paid account gets services that it costs us $X to offer", it's "each paid account subsidizes between 8 and 12 free accounts", depending on if it's paid or premium paid. We've been super lucky that our calculations were pretty much head on at the beginning, and since then, the costs to support each account getting cheaper and the percentage of paid accounts creeping up a bit have more-or-less cancelled out inflation: we don't want to mess with that delicate balancing act by making it possible for people to buy the more popular paid services for a lower cost, because that would mean losing money in the long term.)
no subject
(Before you spend time really thinking things through, though, you should bear in mind that there are a lot of commonly-proposed ideas, such as "let people buy icon slots without having a paid account" and "let people buy access to specific paid features individually for smaller amounts in the long run" that we don't do because it would actually make things worse in the long run. The mailing list messages where I explained the account system and account costs at exhaustive length were, unfortunately, eaten by a case of "we thought we were backing up that old mailing list server and oh fuck, we weren't", but I can do the explanation again if you want it. The short version: we set the prices of paid accounts by what we projected it would cost us to provide the service for everyone, divided by the percentage of accounts that we thought would be willing to pay us, so it's not "each paid account gets services that it costs us $X to offer", it's "each paid account subsidizes between 8 and 12 free accounts", depending on if it's paid or premium paid. We've been super lucky that our calculations were pretty much head on at the beginning, and since then, the costs to support each account getting cheaper and the percentage of paid accounts creeping up a bit have more-or-less cancelled out inflation: we don't want to mess with that delicate balancing act by making it possible for people to buy the more popular paid services for a lower cost, because that would mean losing money in the long term.)