Too much generic comment above!
In flying one should always have an escape route.
Over water, carry a raft.
Over land, a forced landing.
Electrical failure, carry a handheld GPS & radio.
Etc.
Flying IFR/airways, the name of the game is to climb to VMC on top and stay there for the whole route.
So, how thick are the clouds? You can have -3C on the ground, base of 1000ft, stratus tops of 3000ft, and perfect sunshine above that. There isn't going to be enough water in 2000ft of stratus to bring down any normal plane, in the time one can climb up through it (2-3 minutes).
Paradoxically, "UK style IFR" flight (on the IMC Rating, in Class G) is harder in those conditions because it is only in some parts of the UK (Wales, Norfolk) where one can reasonably reliably climb to VMC on top. This causes me to scrap a lot of winter flights - I would have to file them airways which is OK but hardly worth the bother for a 100nm flight.
The business of cloud top data is wrapped up in the whole subject of IFR flight planning.
Winter stratus is generally thinner than in the summer. Summer tops might average 15k; winter tops might average 10k. Often one can see blue sky through holes; that is a giveaway of a thin layer, 2000-3000ft max.
If one took a strict approach (no flight if there is subzero IMC about) there would be no GA IFR flight for half the year, and none at all (deiced planes excepted) on Eurocontrol routes in N Europe because these potentially take you into subzero temps anytime of the year.
Some people won't fly over water.
Some people won't fly over mountains.
Some people won't fly over mountains covered in cloud - I do, and I have a GPS running a topo chart so one could glide into a deep valley, which from FL180 should be possible in most places.
One just has to be a bit clever about it.
The performance drop varies according to type. I have failed to measure an IAS drop with ~ 6mm of rime. Others find a big speed loss. But then I have a TKS prop; I am sure a lot of performance loss blamed on wing icing is actually power loss due to prop ice.