Lumens are measured with a bias around 550 nm or so (More green/yellow range of the spectrum). If you use lumens to describe a different, narrow part of the spectrum (red for example), its not really going to provide equivalent illumination as the same lumens of full spectrum, as the red would have been a small PART of the over all spectrum measured.
So, a 300 L of "normal" light would provide a certain level of illumination....and, depending on the beam angle it was projected with, a certain amount of lux on what you wanted to see.
If the same 300 L light simply filtered out the "not red" light, so ONLY the red part of the spectrum from that original 300 L is sent OTF....the red being sent out would be quite dim in of itself.
If you were to then use your meter to measure that red emission, it would be a VERY low lux level, as it would represent only the red part of the original total 300 L.
If you want to end up with a NEW TOTAL of 300 lumens, with just RED, now you are trying to make red in a scale that mostly counts that green/yellow band, and would require proportionally MASSIVE power to make enough RED to compensate for all the MISSING parts of the spectrum.
Did you specify 300 Lumens of red, with the assumption that it would "look as bright" to you as 300 L of normal light, but just look red?
OR
Did you actually MEAN 300 L with only red light?