I would avoid reading off of a target of saturated colors. Reflected light meters often really do have spectral response problems. I found that my (original) Gossen Ultra Spot produced exposures with variations of up to 4 or 5 stops (!!!) when reading off of colored papers or through colored filters. I was in touch with Gossen designers and, believe it or not, they told me that I should "Never meter off of colored objects at all because I am shooting black & white'"!!!??? Then then told me that I should always meter off of gray cards which, of course, is the equivalent of just taking incident readings and totally ignores the whole point of Zone System readings. (I am not making that up!) I switched to a Zone VI modified meter which gives me nearly perfect densities as claimed. However, I can't claim that meters other than my (original) Gossen meter are that bad. (Gossen finally did admit that they had a problem with the meter and it may have been improved with the second version.) I was surprised to find that my in-camera 35mm meters were far more accurate than the original Ultra Spot.
Anyway, my point is that you should do your testing off of neutral colored targets to be safe.
You should definitely avoid reciprocity during testing for obvious reasons.
I always try to use a target that will produce exposure settings that are fairly normal for my "real" shooting. In fact, the other day I did testing inside myself inside. The daylight in the room was even but low so I used a target that was just about 18% gray. But I avoid white targets and even 18% gray targets if I am testing in "normal" outdoor light so that I can nearly simulate exposure settings that I would normally use outdoors. For example, when I do Zone I (film speed) testing, I use a dark target or I put a medium gray target in deep, even shadow. You should not read off of a white target in bright sunlight and place it on Zone I. If you do, your exposure settings will be very high shutter speeds and/or tiny apertures. After all, when would you place an 18% target on Zone I when it is in brightly lit? The bottom line is to try to have test conditions that closely match your most common actual shooting conditions. If your exposure settings are pretty normal for your real shooting, then you are in the ball park. Again, you should not use targets that produced extreme exposure settings.
An option for indoor testing if daylight entering the room is not enough is to use blue (daylight balanced) photofloods but be sure that they produce uniform readings across the entire target.
When I tested film the other day, the natural light in the room was even but low. (It was raining outside.) To me, that's an ideal testing environment for film speed tests because Zone I readings will be made on similar things under similar conditions. Zone I readings - in real shooting situtions, are taken off of dark things in shadow. But because it was relatively dim in the room, I used an 18% gray card instead of a darker one. As a matter of fact, I have more testing to do today and I think I will do it in my garage using with an 18% gray card. I will place it back in the shadows simulating a real shooting situation as much as possible.
For development tests, I use a white card in daylight, again simulating as close as possible realistic shooting conditions and exposure settings.