A low impedance output to a high impedance input is no problem at all.
Some DI outs are even switchable from mic level to line level (and a hotter output than the 1/4" leads to the conclusion that the XLR out is line level).
Don't think that matching is important. Think of a load (high input impedance: low load, low input impedance: high load) and you won't like to have the load higher than what the out can deliver.
On the other hand a low output impedance means that it can take high load, but a high output impedance means that a high load would let the output signal break down (at least partially).
But I'm not completely sure about the inverse output signal. Usually on mono XLR-1/4" adapters the inverse signal is bridged to ground. If the output was designed well the shortening shouldn't damage the output and it might even be needed. In most if not all cases it might work in that direction without the bridging (just hot to hot and shield/ground to shield/ground). You might want to check the XLR if there is a short from inverse signal to ground.
Hi @DoubleMIDI, yes, please disregard my comment of impedance matching -- a misconception I had that I hopefully cleared after reading info more carefully. I had thought mic-type connections were impedance-matching, but no, they're still impedance-bridging.
I was just thinking that what could benefit the level is a re-amper like this: [Invalid or Expired Link Removed]
My reasoning is that it depends on how the input on the power amp in the MAS speaker is designed. If it expects an input like a regular bass guitar amp, raising the output impedance of the Tonedexter (600 ohms I think) to something more like a bass guitar output (few kohms) may benefit the signal (tone, apparent low-end ??). But anyway... I just barely understand the surface of electronic circuits. I have a reamper and used it successfully at church playing pre-recorded bass tracks on a bass amp.
I think the 6 dB difference of XLR vs. 1/4" comes from: on an XLR connector the input device inverts the inverted signal (pin 3) and sums it to the normal signal (pin 2), thereby achieving the cancelation of any acquired EMI noise. The 6 dB comes from the resulting signal being twice the level, ie. 20 * log( 2 ).
So, maybe a impedance transformer that would do that job of inverting and summing could help. Perhaps something like this (though I'm not sure about the resulting output impedances): Impedance Transformer - Hosa Technology, or this: A15BT Bridging Transformer | Shure Americas