Is it possible to use solar panel(/s) as primary power for an electric bike, HPV or light velomobile (without batteries, mandatory assisted mode only) ? Are there any limitations or obstacles ? Example could an "solar backpack", panel 300Wp 1.6x1m size, weight <20kg help to keep at least reasonable constant speed on a flat road. There could be 5x output power difference depending on sun light, season, etc. (home installation webs) Suppose a single vehicle - no large separate trails, etc.
A human produces 100-300 W (100 W is trivial even when exhausted, 300 W can be produced only up intermediate-length hills). Unless you want the bicycle to be unstable up hills, you want the full 300 W up steep hills. Solar power constant is about 1000 W / square meter, when there are no clouds. Of those, common silicon solar panels produce about 200 W / square meter, the rest is losses. You need more than a square meter of panels and continuous sunny weather to get up hills at any reasonable speed.
If multijunction cells are an option (expensive), you get 400 W / square meter perhaps. So, less than a square meter is enough in sunny weather. You however need to handle cloudy weather somehow.
My advice? Use multijunction cells that track the sun and don't ditch the pedals. The pedals come handy when it's cloudy. I would also advise to not ditch the battery. It helps you up hills, so that even 100-150 W average power produced is enough.
A problem is where to mount the cells and how to ensure they track the sun when you turn. If they don't track the sun, you need more cells.
Also, how to ensure that the solar cells don't cause excessive air resistance is a problem. The mount needs to be very sturdy so that they don't fly away in the wind.
I think you'll find that common backpacks rarely have more than 0.2 square meters of area. Multijunction panels that are 0.2 square meters produce only 80 watts when they are perfectly oriented and there are no clouds. A backpack most of the time is not perfectly oriented, and often there are clouds.
A better approach would be to connect the solar panels to electricity grid (then cheap single-junction silicon cells are enough) to produce hydrogen during the daytime. The hydrogen is then compressed. During the nighttime, when you sleep, the battery of the e-bike is charged using fuel cells powered by the compressed hydrogen. With large enough battery, the system might be enough for riding one day at a time.
You can also refine the grid-connected idea by adding wind power to complement the solar power during windy but non-sunny (cloudy) days.
There is a race, the Sun Trip, for solar-powered bikes (and there have been a few other one-off efforts). Looking at the entrants, there are some that use huge PV roofs to avoid using trailers, but as @Affe points out, those could turn into sails.
Most e-bikes use 250-W motors (if not larger). Depending on the speed you're satisfied with, you could get away with half that power. Weight isn't what will make it hard to maintain a decent speed with these: aerodynamics are vastly more consequential, and a big flat PV array will make it hard to optimize aerodynamics.
Depending on how many hours of the day you're riding the bike, how many hours your PV array is in the sun, and your battery capacity, you may be able to run the bike without ever plugging it into a wall…but you could extend that logic to using a stationary PV array as your power source and running only off batteries—that is, a regular e-bike—when on the road. This might turn out to be more efficient than trying to charge while you ride due to the inefficiencies of having a bike-mounted PV array.
My own take is that solar-powered bikes are possible as interesting experiments, but impractical for day-to-day transportation.