A few references:
indicates (without any supporting evidence) BG speed and a 45 degree bank. They don't say whether that's BG in a turn or S&L, but since 1G BG is all anyone ever reports, that's what I'll assume they mean. In my plane at the GW's I was at, BG is somewhere in the 80 - 85 KIAS region.
is an empirical study that tested success with your parameters - just above stall speed, with different bank angles - 30 degrees and 45 degrees seemed to work the best in these test cases.
is a theoretical analysis, but is backed up with empirical results from testing, and recommends a speed of 5% above the stall speed IN THE TURN.
So we may be saying similar things here, due to the differences between stall speeds in the 45 degree bank turns and at 1G (a factor of about 1.2). My airplane stalls, in the configurations I was testing, at about 62 KIAS. So with a 45 degree bank, the stall speed would be 74.5 KIAS. Now add on the 5% margin recommended in the last reference, and we're at 78 KIAS - almost exactly the 80 KIAS I noted was the best case for us. There's no way MY plane could do a 65 KIAS indicated turn at 45 degrees of bank - I'd be below my indicated stall speed and there's obviously be no margin on top of the stall speed. If we assume that (as is always the case) between min sink speed and stall, the descent rate increases, then it's better to be somewhat above stall, which the 80 KIAS gives me, per the last reference's recommendations.
I think that the canard capability to get right up to stall speed makes this maneuver a lot safer than in a conventional plane, where a 45 degree bank at 5% over stall speed is pretty much begging for a stall/spin accident.