Bicycles are light

You know how you can tell no one with any power really wants to solve the problem?

They won't spend money to see what works.

The easiest historical examples are naval propulsion advances where entire classes of battleship get built to test the operational effectiveness of a new kind of steam engine, but still.  "Does this work? let's find out, here's some budget" is the sure and reliable sign of a sincere desire to solve the problem.  No one in Canada with actual power wants to solve the safe bike travel problem.

The problem with city streets is the fixed width.  Much as I like the mental image, coming by with Really Large hydraulic jacks and pushing all the buildings further apart is not a practical approach.

Car Culture is not an engineering problem; it's a law enforcement problem and a politics problem.  (I have been to a public "talk about the bike lanes" meeting where someone spent forty minutes trying to get a city councillor to commit to increasing the free on-street parking on their adjacent-to-the-proposed-bike-lanes residential street.  They were passionate, committed, and overtly and explicitly of the view that anything involving bicycles was irrelevant, it doesn't matter if people interested in bicycles want to talk, here's a chance to get more parking, that's important and I pay taxes.  Nothing like an engineering problem.)

There is, however, an engineering solution for safe bike travel.

Modern construction methods make it very easy to show up, drill a hole, and drop a (remarkably strong) substantial steel pipe upright in the hole.  There's a lot of stuff under some downtown streets but there's pretty much zero chance a team of engineers can't figure out how and where to put the upright. Modern construction methods make it just as easy to get a lot of truss bridge sections to connect between the pipes that are now suddenly bridge pylons. (There are already mixed-use pathway bridges like this all over the place.)  If you happen to have light rail, you do this over the light rail and use the underside of the bridge sections to support the power wires.  You put a roof over the bridge sections; you put up big cable net guard rails with some spring to them.  You make the whole thing six metres or seven metres wide with a raised "dead" centre section between the lanes so people have some place to fix flats or just get out of the flow of traffic while they're trying to read a map or use their phone.  You pick the post heights to smooth out variation in the ground so there's a relatively continuous and slow change in elevation.  You put in elevated roundabouts over major intersections.  (Yes, really; you can even do a search for "elevated bicycle roundabout" and find examples of existing ones.)

You don't put pedestrians up there.  You might put off-to-the-side bits of path for downtown business that want to put a bicycle cafe on the second or third floor.  You might put in some elevators where there just isn't room for enough ramp.  The dead centre means you can nearly always put in one ramp; there are a lot of places you can put in on and off ramps, down to bits of park or public squares or anywhere you're not dropping the bikes straight into vehicle traffic or pedestrians.  Down into distinct protected bike lanes is fine, and you'd expect to do that out in the city periphery where you have enough width for bike lanes down at grade level.  (Though remember that elevated bike lanes are a solution for the "there are cars turning across my bike lane" problem, and it's fairly tough to make that go away when the roads and the bike routes are at the same grade level.)

While you're at it, you change the laws so building owners are required let bicycles inside; the single largest non-squashing barrier to bike commuting is a complete lack of safe storage.  Many office buildings have ferocious "no plebeian bicycles in the building" policies.  Those need to go.

So -- There's room to put in a LOT of this stuff.  (Every road with an LRT line, just for starters.) You want it to be extensive and continuous; you might want some of it to be fast and some of it to be scenic, but let's not worry about that until there's a lot installed.  It's not hard to imagine a few firms who specialize in assembling and maintaining elevated bike paths, and a resulting number of pretty decent construction and engineering jobs.  (It's not hard to imagine a city department that does this, either.) You design the elevated pathway to be bolt-together so when the day comes and you have to remove some because of major road work, you can come back and unbolt them and carry them away with the same machinery that put them up.  You get a tourist attraction.  You get a business sector where you manufacture prefab bridge sections, pylons, LRT wire hangers, battery-powered snow clearing and surface-cleaning machines, springy cable barrier nets, and so on.  You get an expansion in useful downtown restaurant, cafe, and tourist tchotchke retail space.  You do great things to the official livability index.

And, bicycles are light.  Really light.  Even packed in really tight, you're never going to go over a tonne per metre, counting both lanes (that's four hundred-kilo people per each 2m lane and two more hundred-kilo people in the centre section) and that's just not a challenging truss requirement.  You can likely do this in aluminium (won't rust) and you can order the stuff by the kilometre.

Yes, sure, you get some construction, but it's fast construction; someone is coming in with a drill and a crane for a few days to assemble prefab sections, it's not the deep-pit style of construction.  In terms of dollar per person kilometre of commute, it is WAY cheaper than pretty much everything else.  It gives you a place to stash a lot of cabling in an easy-access simple-maintenance location.  There's no inherent requirement to follow a road, so if you need to bridge a divided highway or a river or railroad, you're going to need some fancier pylons and a longer bridge section.  It gets the bike paths OFF the road.

Anyone in a position of power who actually wanted to solve the bike-safety problem would be trying this.

(The first place you do this in Toronto is Queen's Quay, Bathurst to Parliament.)