fix: getnewaddress - don't blind bech32 addresses

[delta1]

fixes #1245 - getnewaddress will now return an unblinded bech32 address even when -blindedaddresses=1 in config

[apoelstra]

059a4a6 looks fine to me.

Locally the following tests fail

elements-059a4a> feature_confidential_transactions.py               | ✖ Failed  | 1 s
elements-059a4a> feature_sighash_rangeproof.py                      | ✖ Failed  | 2 s
elements-059a4a> rpc_signrawtransaction.py --legacy-wallet          | ✖ Failed  | 6 s

These look like real failures; they happen consistently.

[delta1]

Thanks @apoelstra - fixed

[apoelstra]

ACK 49c3ea3

[roconnor-blockstream]

Now I'm a bit concerned that this change will break existing workflows of people by unblinding their addresses on thm.

[apoelstra]

It may, but I think the old behavior was probably wrong. Not sure how best to handle this.

[psgreco]

I think the old behavior was at least inconsistent, so I'm fine with a comment added to the release notes and this fix making things consistent