I'm not in a position where EtG would seem reasonable, but I can imagine the psychological obstacles which would arise if I was in that position. E.g.:

If you're one of the x-risk-oriented people (like me), rather than, say, global-poverty-oriented, your money wouldn't typically go to people who are much worse off than you, in Africa and elsewhere. It would typically go to support people like AI and generalist researchers, content creators, event organizers, and their support staff—people who are notably better off than you. They spend their days doing work which feels meaningful and enjoyable, often they live (and pay rent!) in the Bay Area, surrounded by fellow EAs and Rationalists, and they enjoy the high social status that the EA community assigns to people who do direct work.

Meanwhile, you spend 8 hours a day doing… well, a job. The people there might be nice enough, but probably not exactly your kind. You're probably working on something that (roughly speaking) doesn't matter. And your future prospects are gloomy: if you really give away a significant portion of your income, rather than saving up, you'll keep toiling as a wage slave for deeecaaadeees before you can afford to retire.

This is indeed something that might make rational sense (if you're somehow particularly ill-equipped for direct work), but it just feels… unfair?

This post gives an excellent description of some challenges of earning to give:

https://80000hours.org/2015/06/why-i-stopped-earning-to-give/

This post is from 2015, but I think the reasoning is still valid. The author stopped earning to give because he 1) performs better working for a cause he believes is important than for the business he used to work for, 2) does not see excellent giving opportunities and could have more impact by doing something else, 3) had different values than his colleagues.

I sort-of earn to give myself and have similar challenges, but I can overcome them and enjoy being with my colleagues even if their values aren't exactly the same - there are always work-related or everyday life topics we can exchange about. I don't agree that there is a lack of giving opportunities, in 2015 it might have been true. Nowadays more charities and cause areas have been evaluated and there are the EA Funds.

[edited to add a little more nuance]

My E2G experience:

In the summer of 2016, I decided to continue self- and university training as a software engineer instead of switching to study biotechnology (I was accepted for two bachelor programs). The main influencers of the decision was my high uncertainty in having an impact in the latter case (relative to a much less risky E2G career), made more salient by my chronic depression.

Having been working as full-time software developer in Berlin since summer '18, I find the job^ de-moralizing, the environment of the "common" people intellectually toxic and my impact in terms of reducing suffering in the world (via E2G in this case) insignificant. I think, dedicated EAs can do better than spending their productive time like this.

If I manage my worsening mental health, I will quite this summer to find a better EA application of my life. (I would like to switch from software engineering to a different field, but due to my Russian citizenship and the official training, I don't yet see how I could get a visa in any promising country.)

^ The company does media monitoring SaaS business. I felt desperate enough to accept the offer.

An answer from personal experience:

It can feel frustrating to be separated from your impact. People in EA have some capacity to resist this, in that we're more likely to give money to help strangers and non-cute animals and people who live in the future, but that doesn't mean we're immune.

Giving money from my basement in Wisconsin while I earned-to-give just didn't feel as good as going to work every day does now. I knew intellectually that I was probably making an impact, but I didn't have any mental image of what the impact looked like, or an especially clear story I could tell myself. Now that I work for CEA, I see the "story" every day, and I've seen dozens of examples of my work being directly linked to impact.

An answer from theory:

It can be hard to trust other people to do good. I'd guess that EA has an unusually high proportion of people who, consciously or not, endorse statements like "if I don't do it, no one will" or "to get something done right, you've got to do it yourself".

Even if it's likely that the person running Organization X is probably better at it than me, I still might have lingering thoughts of "if I could pay myself to run Org X instead of paying them, the world would be better off".