1) The answer is somewhat dependent on your requirement.
Yes tilt at latitude if the panel inclination angle must be fixed, *and* if the desired goal is specifically to maximize average yearly energy collection. However, if the goal is to maximize the minimum power output, which may be the case for a remote sensor powered by PV and the like, then the angle should be much higher to optimize collection for the winter sun (and also to aide in shedding snow or debris cover) at the cost of lost but unnecessary energy in the summer.
Here for example is a PV powered stoplight (supposedly required only in the day) in Houston, Tx, latitude ~30N. I'd guess this panel is mounted 50-60 deg.
If the panel angle can change, then through out the year the angle should be:
[PLAIN]http://www.pvresources.com/images/location/tiltangle.png
2) See NREL's monthly/yearly data http://rredc.nrel.gov/solar/pubs/redbook/" : fixed plate tilted at latitude - 6.5 kWh/M^2/day annual average, minimum 5.9, maximum 6.7. Two axis tracker - 8.9 kWh/M^2/day annual average (370W/M^2 daily average power).
Note that if you have known shading figures (i.e. trees or other obstruction) for your location, for most existing panel designs shading won't translate linearly to reduced electric power output. That is, because of the way in which most all cell designs are wired together, a shading of (say) 25% of a panel's area might cause a reduction in electrical power output in 50% of the PV cells on the panel.