I agree with you and disagree with you at the same time.
(I think your numbers of deaths on the roads ~17,000 is a bit high if you are talking about the UK, although I agree for cyclists, it's about 100 killed each year in the UK. Are you including the seriously injured too in the 17,500 figure? )
The raw figures for road deaths in 2018:
The road safety charity. For safe and healthy journeys
www.brake.org.uk
1784 deaths: 99 cyclists, 777 car occupants, 456 pedestrians, 354 motorcyclists, and I assume the remainder are from coaches, buses and lorries.
So more car occupants die each year on the road than cyclists, from the raw data. But people travel more by car than by bike in the UK. The question is then relative risk and how this is calculated. Do you do this by mile? By number of trips? How do you get a representative estimate? How does it vary by region and urban vs. rural? How does it vary by age of the cyclist? Depending on the model you use and the assumptions you make, you can draw many different conclusions.
Also, by your argument, you would also never walk anywhere, due to the number of pedestrians that get killed each year standing on pavements by cars. So you'd have to go everywhere by car in order to be "safe".
However, their are many other risks associated with driving.
Health is affected by lack of exercise and this can cause early death. Pollution from cars also kills many thousands more each year than from collisions. Not to mention deaths that occur due to climate change brought on by CO2 from cars.
Unfortunately, I don't think anyone has looked at bringing all these factors together into a comparative statistic.
I do agree that 100 cyclists killed each year and about 17,000 seriously injured is far far too high and this is only going to be reduced by driver education, reduction in speed limits, effective enforcement and the creation of a transport infrastructure that is centred around people and not cars (ie walking and cycling).