Reject to run if the backend is still doing IBD

[sr-gi]

Fixes #223

[mariocynicys]

This should detect if we are in IBD while bootstrapping. That's fine.
It doesn't address the issue in #223 though. In #223, it's after bootstrapping is complete and we are in a steady state, bitcoind responds to us slowly but move up the chain quickly (because it was out-of-sync) and prunes blocks we didn't yet receive.

It's not like we can eliminate this case by somehow making bitcoind respond to our RPC quickly, but we can detect the case (we get out-of-sync with bitcoind, i.e. the next block we want is no long with bitcoind) when it finally happens by checking if blocks are actually connected from the output of SpvClient::poll_best_tip.