Base case #1 - If you wire them in series using Vmp you'd have: 20.4v + 20.4v + 16.77v + 16.77v + 17v = 91.34v * (lowest amps = 9.45a) = 863w. This would be 863w/940w (max) = 91.8% of the max.
Base case #2 - If you wire them in parallel you'd have: 9.45a + 9.45a + 11.2a + 11.2a + 11.76a = 53.06a * (lowest voltage = 16.77v) = 890w. This would be 890w/940w (max) = 94.7% of the max.
Then there would be combinations - 1st 2 in parallel ----series---> 2nd 2 in parallel ----- series ----> 5th panel - as an example but the rules should apply - e.g. in series the voltages add up but its the lowest amps that rules and in parallel the amps add up but its the lowest voltage that rules.
In base case #1 you'd be at 91.34v against 100v max of the Victron - likely be OK. In base case #2 you'd be just a bit more efficient (in theory) but you exceed 50a max of the Victron at 53.06a. With 5 in series you'd be subject to loss of power on *any* shading - even 2 x 2 inches kind of thing.
Here's a youtube discussing mixing panels of different types I used as the basis for the above - start at 4:42 for different types:
Personally - 92% vs 95% is no big deal because in real life there are a lot of variables with panels including tilt, shade, clouds, heat etc. However, I would think about shading - for example on my trailer I have each side in it's own string so I can have shade down 1/2 the trailer roof from front to back and still have 1 side producing.