to be honest i don't agree with that wiki article. i have argued about it on reddit too when it was posted there. basically i believe that it is saying "since people are using it wrong, then the method should be considered bad!"
1. Printing is problematic (O.K. I get it that paper is not forever. But let's just agree that I have the archival issues on this front sorted out)
4. Raw Private Keys are Dangerous (I read this section and I do fully understand what they are talking about. So no worries for me on this front.)
simply encrypt everything you want to print and that solves multiple issues including the printer memory, and physical access to the paper containing the keys.
2. Promotes address re-use (I'm talking about sending ONE transaction to this address. No more. Then sweeping it when ready to spend. So I'm calling this a non-issue)
correct call.
3. Encouragement of centralized and outsourced validation (I'm not sure I totally get this one? It might be true that I would use a blockchain explorer to verify the transaction went through. But after that I would not look at that address again. Perhaps I visit the block chain explorer while running a VPN? So I am anonymous while verifying that my paper wallet transactions have indeed landed in their correct addresses.)
it is about "privacy" and that is something most users don't care that much about. for example there are a lot of users already using Electrum so they are already relying on Electrum nodes, one more address is not a big concern to them. if you want privacy then you run your own full node and since paper wallets is something you use once when you are spending, the time rescan takes in your full node is not a big concern either compared to the privacy you gain.
5. Change addresses are not handled which leads to screwups (I understand this. But as long as I sweep the paper wallet into a TREZOR wallet or Core Wallet and spend the coin from the desktop wallet I shouldn't have any problems. Right? I get that if I sweep the coin into the wallet and then uninstall the wallet it won't recover that coin when I use my mnemonic phrase to restore the wallet elsewhere. I would need my paper wallet again.)
the point that this is trying to make is when for instance you have a paper wallet containing 1
BTC and want to sell 0.9
BTC of it. you will be left with 0.1
BTC which if you send back, it will be address reuse and if not you will have to create a new wallet from scratch which is time consuming and hard. (using mnemonic helps solve this)
6. Low Error Correction (I get this concern. But I am totally O.K. with it because I do not intend to have my paper wallets sitting around with the private key printed on them verbatim anyway. I have an entire encryption scheme I intend to use which will eliminate any issues on this front. And don't get me started on why I shouldn't encrypt my private keys. That should be obvious and I'm not buying into the "DON'T DO IT" crowd.)
use BIP38 encryption and create more than one backup of it and you will be fine.
8. Javascript software (I do not fully understand this. I'm open to being educated here)
JavaScript is known to have many bugs and issues. there has been a lot of exploits in it too. and if you look at one of these tools (like bitaddress) history you can see some of these bugs that were found and fixed. generally it is considered unsafe.
9. Browser wallets are bad (I also do not fully understand this. Again... Educate me. Surely there is software out there that can create safe bitcoin public addresses and their matching Private Keys?)
this again is repetition of "users use these tools the wrong way" argument. the "website" is meant for the preview not for usage. the correct way is to download the source code from GitHub and transfer it to a clean offline computer and use it there.