Base Value of +2 seems to be a worst case security: a good start and a bad start are closer (percentage wise). An increase of the decay rate would effect production and food, too. So this seems to be a very radical step. Maybe it would be better to get back to your first idea: increase the cost of early techs?
Maybe I would take too big of an approach and would not get the job done. My "big approach" would consider (at least): game speed, number of habitable planets in the galaxy, techtrading enabled
- Techs are seperated into tiers from early game to super late game (maybe 5)
- Game speed effects techs of each tier the same (today the effect is demolished the farther the game progresses)
- number of habitable planets in the galaxy: no effect to low tier techs and a high cost reduction to late game techs if you play in a small galaxy
- techtrading enabled: no effect to tier 1, scaling up, high effect to super late game techs
- or you may think about techtrading disabled instead - if "enabled" is considered the standard - : you would get a cost reduction (0 at tier 1, scaling up)
The result would be (in case of success...): you could play a game in a small galaxy without tech trading and finish the techtree (in a satisfying manner) and the same would be true for a big galaxy with tech trading enabled.
You could visualize this concept during galaxy setup (a slider, some numbers, a tooltip)