I like the Gossen meters. Good quality silicon blue sensors in something like the Lunasix or the Sixtomat.
You can take the average of the two readings, incident and reflected to bias an incident reading for the reflectivity of the subject. At the end of the day, the camera is only going to take one exposure, the important thing is that its not too short.
Just in case you havent thought about this, before thinking about spot meters vs incident etc, have you thought about testing your processing. Get a simple incident meter if not, take an incident reading and photograph a patch of grass, zones 1 through to 9, (ie 4 stops under to 4 stops over) and add a frame for lens cap on. Process. Look at the negatives. If the step between frames is not correct, the processing time and temp are wrong. If the zone 1 is either too dark or has no detail, the film rating is wrong for the specific meter/lens combination you have used. I think a meter is secondary to this, as suggested above, only one shutter speed and aperture combination is going to be used. You just need to make sure its within 3 stops faster than that required for the shadow detail you want to record. In a normal scene, its likely that an incident reading is going to accomplish this.
Another nice feature of a Sixtomat is that you can take a continuous reading, waving the meter around a bit, and it will record the Subject Brightness Range. This ensures you get shadow detail into the negative.
However, unsuitable processing for the film in question will render all consideration of meter irrelevant.