The paper actually has a nice one-sentence summary:
"The group velocity of light in free space is reduced by controlling the transverse spatial structure of the light beam."
The key word here is "group velocity"
"The group velocity of light in free space is reduced by controlling the transverse spatial structure of the light beam."
The key word here is "group velocity"
Basically this ArXiv article shows that when a light beam has an unusual spatial shape, its group velocity can become measurably slower than light. The group velocity is usually equal to the actual speed of photons, so if it's still the case here, the authors have shown that in certain conditions light can propagate in vacuum slightly slower than it usually does.