Keep in mind that is the constant (average) power. Assuming this is a Q-switched laser with ~10 ns pulse duration, peak power is ~70 kW. (Kind of low as far as lasers go these days, but it's just a laser marker)
I've experimented with using a 60 W (I think*) laser on a small steel bracket, and even with the beam holding on a single point for a minute, it made a barely visible dot that you couldn't feel by running your finger over.
* It was nearly a decade ago, but I looked up the relevant hacker space and unless they changed the model, it was 60 watts.
The bracket was around 1cm by 5cm, and around 1mm thick.
A 60W CO2 laser won't touch steel (or even paper thin aluminum foil), while a 10W diode can cut through it. The type/wavelength of laser matters greatly.
60w back then would be a CO2 LASER - that's 10600nm and that basically bounces off any non-oxidized metal.
The fiber marking LASERs at work are 1064nm, and at a mere 20w output, will absolutely eat away at steel with no problems.
Edit: I should note there are CO2 metal cutting LASERs, but they are at very, very high output powers to overcome that reflectivity barrier. You need 500w 10600nm to cut through what a 30w 1064nm could cut. My 80w CO2 barely cuts through heavy-duty aluminum foil, and in many spots it isn't a full cut. A 20w marking LASER at 1064 would obliterate the foil.
I'm sorry, but there's no way around it, with 5W you can cut metal just as easily as you can cut a tree with a knife. At least with contemporary technology (In the future who knows).