I agree with James K's observation that the main point of a total eclipse is not the darkness but the visibility of the corona.
However, to answer your original question, blocking 99.5% of the Sun* would reduce the daytime illuminance by a factor of 200, roughly from 20,000 lux (full daylight) to 100 lux. Even so, this would not feel very dark, because our eyes have a very wide dynamic range.
At dawn or dusk (in the absence of an eclipse), natural illuminance of 100 lux occurs during civil twilight shortly before sunrise or shortly after sunset. You can still see well. For additional context, this is comparable to a day with very heavy overcast clouds, or to the indoor lighting in an office building.
So the change in general brightness during a 99.5% eclipse is noticeable, but not dramatic. Even if the eclipse is covered by clouds, there would be a much bigger perceived change in brightness where the eclipse is total. And if the eclipse can be viewed clearly, then of course there is no comparison.
* A technicality: The most commonly quoted measure of a partial eclipse, the "magnitude", is the maximum fraction of the Sun's apparent diameter (not area) that is covered by the Moon. So for a precise answer, you would need to ensure which number you read. I will assume here that 99.5% is the fraction of area, known as the "obscuration". The issue does not affect the qualitative conclusions.